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

Error in Yocto Tensorflow integration #2

Open
gija92 opened this issue Aug 29, 2019 · 2 comments
Open

Error in Yocto Tensorflow integration #2

gija92 opened this issue Aug 29, 2019 · 2 comments

Comments

@gija92
Copy link

gija92 commented Aug 29, 2019

Hello,

I would like to build an Yocto Image which inherits the recipies Tensorflow-Lite and ARMNN. But I get following errors:

Armnn:
asd

Tensorflow-lite:
fff

I'm working on an ARM-based machine (Renesas R-Car H3, aarch64).
Hope anyone can help me out. Thanks for every advice

@gija92 gija92 changed the title Error in tensorflow build Error in Yocto Tensorflow integration Aug 29, 2019
@cevincent
Copy link
Contributor

Hi Gija,

Instead of screenshots, could you provide us with all the logs in text format?
These recipes were tested only on ARMv7, which probably explains these errors. You can try the meta-layer provided by Renesas:
https://github.com/renesas-rz/meta-renesas-ai

Keep in mind that if you use this layer, the version of tensorflow lite and ARM NN are quite old, but otherwise they should probably work out of the box with your processor.

Thanks!

@gija92
Copy link
Author

gija92 commented Sep 9, 2019

Hi Vincent,

Logfiles:
ErrorLog_ArmComputeLibrary.txt
ErrorLog_TensorflowLite.txt

In my thought Tensorflow and ARMNN should run on the Cortex R-Core (ARMv7) which is included on the board. Is there a way to "include" the R-CPU into the Yocto Build or is an integration of these libraries better on the A-CPUs regarding performance?

Thank you

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants