Open Models Almanac
Methodsoptimizationtraining parallelism

single mention · optimization

Combined context, tensor, and expert parallel training

The training configuration combines context, tensor, and expert parallelism, with the cited degrees specifying one implementation.

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.

Filed under optimization › training parallelism