Open Models Almanac
Methodstraining objectiveauxiliary loss

single source · training objective

Domain-specific KL regularization strength

Varies the strength of KL regularization across domains.

The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See auxiliary loss for the methods filed beside it.

Also called Domain-adaptive KL regularization strength, varying strengths of KL regularization.

Filed under training objective › auxiliary loss