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.

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

Logging suggestions #75

Closed
awgu opened this issue Feb 23, 2024 · 2 comments
Closed

Logging suggestions #75

awgu opened this issue Feb 23, 2024 · 2 comments

Comments

@awgu
Copy link
Contributor

awgu commented Feb 23, 2024

  1. Should we log the absolute GB, not only percentage? It might be more helpful as a language to compare across different GPU types.
  2. Should we log WPS?
@lessw2020
Copy link
Contributor

imo, yes and yes.
tok per sec (instead of wps) and mfu are planned, to build on point #2.
could also add total_tokens and batches_per_sec as additional items.

@awgu
Copy link
Contributor Author

awgu commented Mar 5, 2024

Closing in favor of #3

@awgu awgu closed this as completed Mar 5, 2024
payoto pushed a commit to graphcore-research/torchtitan-fork that referenced this issue Feb 7, 2025
@ganeshk-gc
Sweeps (pytorch#75)
6ea8f53 
* add zip
* add e2e test, add types
* add test_titan_sweep.sh  back, used as default command in script
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@awgu @lessw2020