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

Exploit symmetry of covariance matrices for faster & more stable diag… #35571

Open
wants to merge 1 commit into
base: main
Choose a base branch
from

Conversation

bohrium
Copy link

@bohrium bohrium commented Jan 8, 2025

Fixes # 34570

see #34570

@Rocketknight1

…onalization during initialization of added embeddings
@bohrium bohrium requested a review from Rocketknight1 as a code owner January 8, 2025 23:58
@Rocketknight1
Copy link
Member

Hi @bohrium, thanks for the fix!

However, there's one other thing I realized, and I'm sorry I forgot about this over the Christmas break. Another user pointed out in #35075 that we don't need to explicitly generated the eigenvals at all, because we can just use a constraint check from torch.distributions. They never got around to making the PR, though - maybe you could change this PR to use the constraint check instead? That should be even faster than eigvalsh.

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

Successfully merging this pull request may close these issues.

2 participants