Skip to content

Non-matching pairs are scored way too high, well above threshold #3

@rjurney

Description

@rjurney

This makes the model unusable atm, which is confusing, given its high performance scores... could it be overfitting that much? Suspect an error in the training setup.

Metadata

Metadata

Assignees

Labels

No labels
No labels

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions