Open Models Almanac
Methodstraining objectivedistillation objective

single mention · training objective

Combined KL-divergence and cross-entropy distillation loss

This objective combines KL distillation and cross-entropy with separately weighted terms.

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

Filed under training objective › distillation objective