dumbnet 0.1.0

a [no_std] neural network library
Documentation 
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

use dumbnet::{
	activation::Sigmoid,
	layers::{Layer, OutputLayer},
};

use generic_array::typenum;

fn main() {
	let mut last = OutputLayer::<Sigmoid, typenum::U1, typenum::U2>::new();
	let inputs = vec![
		([0., 0.].into(), [0.].into()),
		([0., 1.].into(), [1.].into()),
		([1., 0.].into(), [1.].into()),
		([1., 1.].into(), [1.].into()),
	];
	last.teach(inputs.clone().into_iter(), 1000, |_, _| {});

	for (input, output) in &inputs {
		let result = last.calculate(&input);
		println!(
			"trained result of {:?} is {} should be {}",
			input, result[0], output[0]
		);
	}
}