Open Models Almanac
Methodsinference & servinginference quantization

single mention · inference & serving

Per-block FP8 quantization for attention Q/K/V projections

FP8 quantization of attention query, key, and value projections using per-block scales.

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.

Filed under inference & serving › inference quantization