Open Models Almanac
Methodsoptimizationtraining parallelism

single mention · optimization

Model parallelism

Model parallelism uses multiple GPUs to reduce per-GPU memory requirements; the evidence does not identify a more specific method.

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 Multi-GPU model parallelism.

Filed under optimization › training parallelism