oa::FnMatrix::quantize

Float32 storage into one native OA Q4 or Q8 weight value. The returned value retains the source shape; its physical planes remain private.

FnC++ public headers

Function Documentation

oa::QuantMatrix oa::FnMatrix::quantize(
	const oa::Matrix & inInput,
	oa::Quantization inQuantization
)

Float32 storage into one native OA Q4 or Q8 weight value. The returned value retains the source shape; its physical planes remain private.

inInput

const oa::Matrix &

—

inQuantization

oa::Quantization

—

oa::QuantMatrix

The declared return value.