Glossary term
Glossary term
Foundations
A metric representing a model's loss against the test set. When building a model, you typically try to minimize test loss. That's because a low test loss is a stronger quality signal than a low training loss or low validation loss.
A large gap between test loss and training loss or validation loss sometimes suggests that you need to increase the regularization rate.
Created for this library
An ML team reports test loss alongside training and validation loss at every release review to detect overfitting.
A research team monitors test loss across model variants to compare generalization on the same held-out set.
An ML platform team requires reporting test loss alongside business metrics so reviewers see both technical and product perspectives.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License