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

Additional training data and model retrain for the segmentation #1200

Merged
merged 5 commits into from
Dec 11, 2024

Conversation

lfoppiano
Copy link
Collaborator

@lfoppiano lfoppiano commented Nov 13, 2024

This should fix #984 #1198 #1187 #1155

The commit is mine, i use my credentials but with the wrong email/name

@lfoppiano lfoppiano changed the title New training data for the segmentation model Additional training data for the segmentation model Nov 13, 2024
@lfoppiano lfoppiano added this to the 0.8.2 milestone Nov 21, 2024
@lfoppiano lfoppiano force-pushed the new-training-data-das branch from d14eb11 to 7fbbdaf Compare November 22, 2024 08:50
@coveralls
Copy link

coveralls commented Nov 22, 2024

Coverage Status

coverage: 40.706% (-0.05%) from 40.755%
when pulling 1e68679 on new-training-data-das
into f4d6245 on master.

@lfoppiano
Copy link
Collaborator Author

Here the deltas for the end 2 end evaluation.

TLDR: very small differences overall (as it should be), DAS recognition has increased from 0.08 to 2.5% in F1 in different datasets.

PMC

Header

changes within -0.1 and + 0.3 F1 scores

Citation

changes within -0.1 and +0.1 F1 scores

Fulltext

  • changes within -0.1 and +0.1 F1 scores
  • Document ratio avail_stmt: N/A for PMC

BiorXiv

Header

changes within -0.1 and + 0.3 F1 scores

Citation

changes within -0.1 and +0.1 F1 scores

Fulltext

  • changes within -0.1 and +0.1 F1 scores
  • Document ratio avail_stmt: F1 +0.4

eLife

Header

changes within -0.1 and + 0.3 F1 scores

Citation

changes within -0.1 and +0.1 F1 scores

Fulltext

  • changes within -0.1 and +0.1 F1 scores
  • Document ratio avail_stmt: F1 +2.5%

PLOS

Header

changes within -0.1 and + 0.3 F1 scores

Citation

changes within -0.1 and +0.1 F1 scores

Fulltext

  • changes within -0.1 and +0.1 F1 scores
  • Document ratio avail_stmt: F1 +0.08%

@lfoppiano lfoppiano marked this pull request as ready for review November 25, 2024 14:40
@lfoppiano lfoppiano requested a review from kermitt2 November 25, 2024 14:40
@lfoppiano lfoppiano merged commit 062e165 into master Dec 11, 2024
7 of 10 checks passed
@lfoppiano lfoppiano deleted the new-training-data-das branch December 11, 2024 14:05
@lfoppiano lfoppiano changed the title Additional training data for the segmentation model Additional training data and model retrain for the segmentation Dec 22, 2024
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.

Incomplete funding statement extraction
2 participants