Open Models Almanac
Methodsinference & servinginference kernel

single mention · inference & serving

Shape-aware kernel dispatch for Hyper-Connection

Selects the appropriate Hyper-Connection execution path for low-latency decode or large-scale prefill.

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

Also called Shape-aware dispatch.

Filed under inference & serving › inference kernel