Open lizgzil opened 9 months ago
Various changes to improve the existing models
Fixes # (issue)
Note on the confidence threshold: After experimenting using the evaluation dataset I found that 0.4 gave good results on all labels.
notebooks/
pre-commit
dev
README
Description
Various changes to improve the existing models
Fixes # (issue)
Instructions for Reviewer
Note on the confidence threshold: After experimenting using the evaluation dataset I found that 0.4 gave good results on all labels.
Checklist:
notebooks/
pre-commit
and addressed any issues not automatically fixeddev
README
s