Multi-SupConLoss #667
Open
Multi-SupConLoss #667
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Adding details to NTXentLoss documentation
Add cross-batch memory for multi-supcon Add test cases
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
I implemented the Multi-SupConLoss as mentioned in issue #658.
However, this implementation rewrote a lot of functions because many functions conflicted with the multi-label task such as check_shapes and get_matches_and_diffs.
Moreover, it seems that the cross-batch memory does not support the Multi-SupConLoss. I found that to support the Multi-SupConLoss I need to modify a bunch of code which will potentially affect other well-developed functions.
So, I tried to add the function I needed to multilabel_supcon_loss.py and didn't make any changes to other well-developed functions.
I also tested the use-case and compared the result with repo. The result looks good except the cases ran on Float16.