Skip to content

Faster inference by caching the KV projections - #15

Merged
rushin682 merged 2 commits into
peng-lab:mainfrom
Harry25R:perf/inference-speedup-v2
Sep 9, 2026
Merged

rushin682 merged 2 commits into
peng-lab:mainfrom
Harry25R:perf/inference-speedup-v2