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

What to use as replacement for tf_cnn_benchmark in the official tensorflow models #535

Closed
rahulbatra85 opened this issue May 2, 2023 · 2 comments

Comments

@rahulbatra85
Copy link

I see a benchmark branch in Tensorflow official models. Is this suppose to tf_cnn_benchmark replacement?
https://github.com/tensorflow/models/tree/benchmark

Also, in an issue in this repo it was mentioned to use officials/vision models as alternative
#524 (comment)

Which one to use?

Thanks!

@tfboyd
Copy link
Member

tfboyd commented May 2, 2023

Things have changed over the years. For official models that also train to the correct accuracy and are maintained models/official is the way to go/. https://github.com/tensorflow/models/tree/master/official. I think it was 4 years ago, when I also moved on from tf_cnn_benchmarks and really pushed models/official because I felt perf code wasn't as interesting as really nice core models that also preformed out of the box with minimal Flags.

The other path are the MLPerf models; but not everyone publishes every model on every platform. I have not been involved in a long time; but MLPerf seems more hardware focused.

I did really enjoy all the fun we had with tf_cnn_benchmarks. Good luck finding what you need.

@tfboyd tfboyd closed this as completed May 2, 2023
@rahulbatra85
Copy link
Author

ok, thanks!

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