Open Models Almanac
Methodsoptimizationtraining runtime

single mention · optimization

Post-iteration NVMe offloading of training states

Offloads model weights and optimizer states to NVMe after a training iteration finishes.

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 offload training states ... to NVMe.

Filed under optimization › training runtime