single mention · inference & serving
FP16 cache storage with stochastic rounding
SSM cache storage in FP16 with stochastic rounding.
The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See inference quantization for the methods filed beside it.
Also called FP16 Cache with Stochastic Rounding.
Filed under inference & serving › inference quantization