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.

Sign up for GitHub

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

Jump to bottom

[QST] how to use example 67? #2060

Closed

ginowu opened this issue Jan 25, 2025 · 0 comments

Labels

? - Needs Triage question

ginowu commented Jan 25, 2025

For the recent example 67, thanks for your great work. I have some questions about the usage in real cases, could you share your thoughts:

One block tile share one pair of scale_A and scale_B, then how could user generate these scales values? Should users need to use together with next version of Transformer Engine?
For abs_max_D, it's just one float value, but as I understand, it would be used for calculating scale_A of downstream GEMM, then it should be multiple values instead of one?

ginowu added ? - Needs Triage question labels

ginowu closed this as completed

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment