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

Refactor contrastive loss #35

Merged
merged 4 commits into from
Dec 3, 2024
Merged

Refactor contrastive loss #35

merged 4 commits into from
Dec 3, 2024

Conversation

fcogidi
Copy link
Collaborator

@fcogidi fcogidi commented Dec 3, 2024

PR Type

Refactor.

Short Description

The main change in this PR is the refactoring of CLIPLoss to ContrastiveLoss. This new loss generalizes the CLIP loss better than the previous implementation (especially in distributed setting) and also includes a modality alignment loss.

Tests Added

None.

@fcogidi fcogidi added the refactor Change exisiting code while maintaining the same functionality label Dec 3, 2024
@fcogidi fcogidi self-assigned this Dec 3, 2024
@fcogidi fcogidi merged commit 3dea00f into main Dec 3, 2024
4 checks passed
@fcogidi fcogidi deleted the fco/reproducibility_study branch December 3, 2024 22:19
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
refactor Change exisiting code while maintaining the same functionality
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant