-
Notifications
You must be signed in to change notification settings - Fork 264
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
add expandable segments var #775
Conversation
…unami enumeration (#764) * adding new notebook for using fairchem models with NEBs * adding md tutorials * blocking code cells that arent needed or take too long
Codecov ReportAll modified and coverable lines are covered by tests ✅
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM.
Can you confirm that there is no increase in memory usage for training smaller models? If that is the case, feel free to merge.
yep: checked here https://fairwandb.org/fairchem/fm_testing/reports/Expandable-segments-on-31M-model--Vmlldzo0MzMxMg |
Expandable segments is a new cuda feature that helps with memory fragmentation during frequent allocations (ie: in the case of variable batch sizes). see https://pytorch.org/docs/stable/notes/cuda.html.
This report shows we can get a memory savings of 4GB for a 0.5B model for example: https://fairwandb.org/fairchem/fm_testing/reports/Memory-Savings-for-EQV2--Vmlldzo0MzEwMA