Open Models Almanac
Methodsoptimizationtraining parallelism

single mention · optimization

Data and expert parallelism for calibration

Data-parallel and expert-parallel dimensions are combined for calibration when pipeline parallelism would leave only one GPU active at a time.

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

Also called DP× EP.

Filed under optimization › training parallelism