1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
|
#[cfg(feature = "mkl")]
extern crate intel_mkl_src;
use anyhow::Result;
use candle::{Device, Tensor};
use candle_nn::LayerNorm;
#[test]
fn layer_norm() -> Result<()> {
let device = &Device::Cpu;
let w = Tensor::new(&[3f32], device)?;
let b = Tensor::new(&[0.5f32], device)?;
let ln = LayerNorm::new(w, b, 1e-8);
let two = Tensor::new(&[[[2f32]]], device)?;
let res = ln.forward(&two)?.flatten_all()?;
assert_eq!(res.to_vec1::<f32>()?, [0.5f32]);
let inp = Tensor::new(&[[[4f32, 0f32]]], device)?;
let res = ln.forward(&inp)?;
assert_eq!(res.to_vec3::<f32>()?, [[[3.5f32, -2.5]]]);
let inp = Tensor::new(&[[[1f32, 2., 3.], [4., 5., 6.], [9., 8., 7.]]], device)?;
let res = ln.forward(&inp)?;
assert_eq!(
res.to_vec3::<f32>()?,
[[
[-3.1742344, 0.5, 4.1742344],
[-3.1742344, 0.5, 4.1742344],
[4.1742344, 0.5, -3.1742344]
]]
);
let mean = (res.sum_keepdim(2)? / 3.0)?;
// The average value should be `b`.
assert_eq!(mean.to_vec3::<f32>()?, [[[0.5], [0.5], [0.5]]]);
let std = (res.broadcast_sub(&mean)?.sqr()?.sum_keepdim(2)?.sqrt()? / 3.0)?;
// The standard deviation should be sqrt(`w`).
assert_eq!(
std.to_vec3::<f32>()?,
[[[1.7320508], [1.7320508], [1.7320508]]]
);
Ok(())
}
|
