Convert a double-precision floating-point number to the nearest half-precision floating-point number. - View it on GitHub