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

ER-ACE criterion simplification #1313

Merged
merged 1 commit into from Mar 6, 2023
Merged

Conversation

AlbinSou
Copy link
Collaborator

@AlbinSou AlbinSou commented Mar 6, 2023

As mentioned in continual learning baselines repo, I simplified the ER-ACE criterion so that it does not track the old and current classes anymore but only considers classes in the current batch as current classes. That way the method does not have to be aware of the stream structure, as it is done in the paper. The results are not changed drastically, sometimes better sometimes worst but not a big difference overall.

@coveralls
Copy link

Pull Request Test Coverage Report for Build 4342486641

  • 4 of 4 (100.0%) changed or added relevant lines in 1 file are covered.
  • 1 unchanged line in 1 file lost coverage.
  • Overall coverage increased (+0.008%) to 73.78%

Files with Coverage Reduction New Missed Lines %
avalanche/training/supervised/er_ace.py 1 60.4%
Totals Coverage Status
Change from base Build 4164094425: 0.008%
Covered Lines: 13850
Relevant Lines: 18772

💛 - Coveralls

@AntonioCarta
Copy link
Collaborator

Thanks

@AntonioCarta AntonioCarta merged commit 5296b7f into ContinualAI:master Mar 6, 2023
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.

None yet

3 participants