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

MACs vs. Flops #7

Closed
mjjung0925 opened this issue Sep 30, 2019 · 2 comments
Closed

MACs vs. Flops #7

mjjung0925 opened this issue Sep 30, 2019 · 2 comments

Comments

@mjjung0925
Copy link

Hello, I have a question about the way you counted FLOPs.
According to your code prune/utils.py, you used thop.
However, the code is recently corrected that it actually measured MACs rather than FLOPs.
(Please refer to Lyken17/pytorch-OpCounter#37
and Lyken17/pytorch-OpCounter@41fd65c)

In this case, are the FLOPs values in your paper still valid?

Thank you.
Jung

@youzhonghui
Copy link
Owner

Thank you very much. I am traveling these days. I will look into it in detail later.

@youzhonghui
Copy link
Owner

According to the following description of MACs and FLOPs differences, roughly MACs = 2 * FLOPs, which would not affect the FLOPs pruning ratio that reported in our paper.
sovrasov/flops-counter.pytorch#16 (comment)

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