Open Models Almanac
Methodspost-trainingsupervised fine-tuning

single mention · post-training

Supervised fine-tuning at 256k sequence length

Runs SFT with a 256k sequence length.

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.

Filed under post-training › supervised fine-tuning