You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tvm.apache.org by "gugudeshubao (via GitHub)" <gi...@apache.org> on 2025/07/25 11:35:51 UTC
[I] [Tracking Issue] 0.21.dev tag do not have c++ runtime inference demo [tvm]
gugudeshubao opened a new issue, #18164:
URL: https://github.com/apache/tvm/issues/18164
I want use tvm on automatic drive。 all the code is write in cpp 17,so i need use c++ to run the tvm export so。
0.21.dev tag do not have c++ runtime inference demo。 is there have some docs or example？I read all the file in doc and tests dir ，i not find usefull infomation。
thanks。
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@tvm.apache.org.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [I] [Tracking Issue] 0.21.dev tag do not have c++ runtime inference demo [tvm]
Posted by "YouCaiJun98 (via GitHub)" <gi...@apache.org>.
YouCaiJun98 commented on issue #18164:
URL: https://github.com/apache/tvm/issues/18164#issuecomment-3162969416
> > I made a small example project showing how to run TVM Relax module with C++: https://github.com/mshr-h/tvm-relax-cpp-example
> > Hope it helps.
>
> Hi, I wonder if it would be convenient for you to also provide a Python demo. I've been going through the docs for a couple of days and found that some parts are somewhat out-of-date, which has caused difficulties in getting started. For instance, I tried exporting the optimized model and reloading to examine it. However, I find the api provided in the doc [Introduction to Module Serialization](https://tvm.apache.org/docs/arch/introduction_to_module_serialization.html) is no longer available. It seems that the 'tvm.runtime.load' function has been removed. I also searched the keywords such as 'save', 'export', 'load' in the documents, but unfortunately didn't find the information I needed. Many thanks in advance :)
Additionally, most of the blogs I could find are based on the 'relay' module, which I reckon is an older module that has been restructured in the latest version. Could you provide a document detailing the interface changes? Thanks!
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@tvm.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [I] [Tracking Issue] 0.21.dev tag do not have c++ runtime inference demo [tvm]
Posted by "gugudeshubao (via GitHub)" <gi...@apache.org>.
gugudeshubao commented on issue #18164:
URL: https://github.com/apache/tvm/issues/18164#issuecomment-3163220757
thanks all. i'll test it on my env。
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@tvm.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [I] [Tracking Issue] 0.21.dev tag do not have c++ runtime inference demo [tvm]
Posted by "gugudeshubao (via GitHub)" <gi...@apache.org>.
gugudeshubao commented on issue #18164:
URL: https://github.com/apache/tvm/issues/18164#issuecomment-3163253510
> I made a small example project showing how to run TVM Relax module with C++: https://github.com/mshr-h/tvm-relax-cpp-example
>
> Hope it helps.
I find in tvm0.21 version no class PackedFunc。Module.GetFunction() return class ffi::Function.
Can you try in 0.21dev.
thanks.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@tvm.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [I] [Tracking Issue] 0.21.dev tag do not have c++ runtime inference demo [tvm]
Posted by "mshr-h (via GitHub)" <gi...@apache.org>.
mshr-h commented on issue #18164:
URL: https://github.com/apache/tvm/issues/18164#issuecomment-3170220799
Updated. Should work with the latest tvm ffi system.
https://github.com/mshr-h/tvm-relax-cpp-example
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@tvm.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [I] [Tracking Issue] 0.21.dev tag do not have c++ runtime inference demo [tvm]
Posted by "mshr-h (via GitHub)" <gi...@apache.org>.
mshr-h commented on issue #18164:
URL: https://github.com/apache/tvm/issues/18164#issuecomment-3146545163
I made a small example project showing how to run TVM Relax module with C++:
https://github.com/mshr-h/tvm-relax-cpp-example
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@tvm.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [I] [Tracking Issue] 0.21.dev tag do not have c++ runtime inference demo [tvm]
Posted by "YouCaiJun98 (via GitHub)" <gi...@apache.org>.
YouCaiJun98 commented on issue #18164:
URL: https://github.com/apache/tvm/issues/18164#issuecomment-3162948365
> I made a small example project showing how to run TVM Relax module with C++: https://github.com/mshr-h/tvm-relax-cpp-example
>
> Hope it helps.
Hi, I wonder if it would be convenient for you to also provide a Python demo. I've been going through the docs for a couple of days and found that some parts are somewhat out-of-date, which has caused difficulties in getting started. For instance, I tried exporting the optimized model and reloading to examine it. However, I find the api provided in the doc [Introduction to Module Serialization](https://tvm.apache.org/docs/arch/introduction_to_module_serialization.html) is no longer available. It seems that the 'tvm.runtime.load' function has been removed. I also searched the keywords such as 'save', 'export', 'load' in the documents, but unfortunately didn't find the information I needed.
Many thanks in advance :)
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@tvm.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [I] [Tracking Issue] 0.21.dev tag do not have c++ runtime inference demo [tvm]
Posted by "thishome (via GitHub)" <gi...@apache.org>.
thishome commented on issue #18164:
URL: https://github.com/apache/tvm/issues/18164#issuecomment-4001497163
hi @mshr-h , have you tried a cpp demo on cuda gpu? I tried your demo with little modification and but failed. below is my change in py and cpp files
<img width="352" height="189" alt="Image" src="https://github.com/user-attachments/assets/a40b3f5f-68db-4206-b063-59510590f127" />
[temp.cpp](https://github.com/user-attachments/files/25754681/temp.cpp) please give me some advice, thank you！
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: discuss-archive-unsubscribe@tvm.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: discuss-archive-unsubscribe@tvm.apache.org
For additional commands, e-mail: discuss-archive-help@tvm.apache.org
Re: [I] [Tracking Issue] 0.21.dev tag do not have c++ runtime inference demo [tvm]
Posted by "thishome (via GitHub)" <gi...@apache.org>.
thishome commented on issue #18164:
URL: https://github.com/apache/tvm/issues/18164#issuecomment-4002554795
https://github.com/thishome/tvm-relax-cpp-example a success version for gpu demo
> hi [@mshr-h](https://github.com/mshr-h) , have you tried a cpp demo on cuda gpu? I tried your demo with little modification and but failed. below is my change in py and cpp files
>
> <img alt="Image" width="352" height="189" src="https://private-user-images.githubusercontent.com/15165355/558398977-a40b3f5f-68db-4206-b063-59510590f127.png?jwt=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoicmF3LmdpdGh1YnVzZXJjb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3NzI2OTE5OTgsIm5iZiI6MTc3MjY5MTY5OCwicGF0aCI6Ii8xNTE2NTM1NS81NTgzOTg5NzctYTQwYjNmNWYtNjhkYi00MjA2LWIwNjMtNTk1MTA1OTBmMTI3LnBuZz9YLUFtei1BbGdvcml0aG09QVdTNC1ITUFDLVNIQTI1NiZYLUFtei1DcmVkZW50aWFsPUFLSUFWQ09EWUxTQTUzUFFLNFpBJTJGMjAyNjAzMDUlMkZ1cy1lYXN0LTElMkZzMyUyRmF3czRfcmVxdWVzdCZYLUFtei1EYXRlPTIwMjYwMzA1VDA2MjEzOFomWC1BbXotRXhwaXJlcz0zMDAmWC1BbXotU2lnbmF0dXJlPTY2ZjFlMDBiNTliNmEyZDYwYWRmMjA2MDUwY2VlZTExNTRlMGU5YzQ3YjI1ODBiN2JmMWQ5M2RkYmQ4ZDAwNDYmWC1BbXotU2lnbmVkSGVhZGVycz1ob3N0In0.86Zx0WEWFOo_uG70uWlvxVBpHSRqsQeZFJviQXbA8-Q">
> [temp.cpp](https://github.com/user-attachments/files/25754681/temp.cpp) please give me some advice, thank you！
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: discuss-archive-unsubscribe@tvm.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: discuss-archive-unsubscribe@tvm.apache.org
For additional commands, e-mail: discuss-archive-help@tvm.apache.org