Open Models Almanac
Methodsoptimizationtraining runtime

single mention · optimization

Checkpointing every 250 training steps

Saves a checkpoint at fixed intervals of 250 training steps.

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

Also called checkpoint every 250 training steps.

Filed under optimization › training runtime