Open Models Almanac
Methodsinference & servinginference quantization

single mention · inference & serving

Embedding and KV-cache quantization

A compression approach focused on the vocabulary embeddings and short-term memory, without a specified quantization format.

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 Embedding and KV cache optimization.

Filed under inference & serving › inference quantization