pub fn f16round(x: f64) -> f64
The Math.f16round() function returns the nearest 16-bit half precision float representation of a number.
Math.f16round()
MDN documentation