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

Better small model support #28

Merged
merged 1 commit into from
Dec 22, 2022
Merged

Better small model support #28

merged 1 commit into from
Dec 22, 2022

Conversation

jimgao1
Copy link
Contributor

@jimgao1 jimgao1 commented Dec 20, 2022

This PR improves support for profiling smaller models by better exploring the range of potential batch sizes to build the Skyline performance model.

@anurlybayev
Copy link
Contributor

How would I test this?

@jimgao1
Copy link
Contributor Author

jimgao1 commented Dec 20, 2022

How would I test this?

Try a small model (e.g. resnet34) with small images (e.g. CIFAR10) as input. We would expect a larger range of batch sizes being tested than before. (i.e. [64,1064,2064] instead of [64,86,104]).

Copy link
Contributor

@anurlybayev anurlybayev left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I didn't test this. But just looking at the changes, it seems that it's all about coefficients.

@jimgao1 jimgao1 merged commit 3839124 into main Dec 22, 2022
@jimgao1 jimgao1 deleted the dec20-small-models branch December 22, 2022 18:19
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

Successfully merging this pull request may close these issues.

2 participants