Open Models Almanac
Methodsoptimizationtraining runtime

single mention · optimization

Caching the distributed checkpoint save plan

Caches a distributed save plan so that it is computed once rather than on every checkpoint save.

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 cache the distributed save plan.

Filed under optimization › training runtime