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
44
45
46
47
48
49
50
51
52
53
54
55
56
mod cpu_kernel;
#[cfg(feature = "cuda")]
mod cuda_kernel;
use super::ops::{try_unary_op, UnaryKernel};
use crate::{shapes::*, tensor::*};
#[repr(C)]
#[derive(Debug, Default, Copy, Clone)]
pub struct ReLUKernelOp;
pub fn relu<S: Shape, E: Dtype, D: UnaryKernel<ReLUKernelOp, E>, T: Tape<E, D>>(
t: Tensor<S, E, D, T>,
) -> Tensor<S, E, D, T> {
t.relu()
}
impl<S: Shape, E: Dtype, D: UnaryKernel<ReLUKernelOp, E>, T: Tape<E, D>> Tensor<S, E, D, T> {
pub fn relu(self) -> Self {
self.try_relu().unwrap()
}
pub fn try_relu(self) -> Result<Self, D::Err> {
try_unary_op(ReLUKernelOp, self)
}
}
#[cfg(test)]
mod tests {
use crate::{tensor::*, tensor_ops::*, tests::*};
#[test]
fn test_relu() {
let dev: TestDevice = Default::default();
let x: Tensor<_, TestDtype, _> = dev.tensor([-2.0, -1.0, 0.0, 1.0, 2.0]);
let r = x.leaky_trace().relu();
assert_eq!(r.array(), [0.0, 0.0, 0.0, 1.0, 2.0]);
let g = r.exp().mean().backward();
assert_close(&g.get(&x).array(), &[0.0, 0.0, 0.0, 0.54365635, 1.4778112]);
}
}