BIBRA should have a way to evaluate how well backend projects work on datasets.
The implementation could be accessed via CLI command, for example bibra eval <project-id> <path/to/file>. There is an evaluation script in the FinGreyLit repository that can probably be used as a starting point.
BIBRA should have a way to evaluate how well backend projects work on datasets.
The implementation could be accessed via CLI command, for example
bibra eval <project-id> <path/to/file>. There is an evaluation script in the FinGreyLit repository that can probably be used as a starting point.