Open Models Almanac
Methodsinference & servinginference scheduling

single mention · inference & serving

Workload-aware routed-expert GEMM scheduling

Adapts routed-expert GEMM scheduling parameters to the current token distribution before launch.

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

Also called workload-aware scheduler.

Filed under inference & serving › inference scheduling