Open Models Almanac
Methodspost-trainingsupervised fine-tuning

single mention · post-training

Last-layer and token-embedding fine-tuning

Updates only the final transformer layer and token embedding matrix.

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

Also called last layer only.

Filed under post-training › supervised fine-tuning