Jsevillamol comments on How to measure FLOP/​s for Neural Networks empirically?

[Jsevillamol]

Following up on this—the aforementioned second blog post is now up!