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
57
58
59
60
61
62
63
64
// use crate::layers::*;
// use crate::model::*;
// use crate::optimizer::Optimizer;
// use crate::save_load;
// use crate::spiral::*;
use std::time::Instant;
pub fn benchmark() {
let start = Instant::now();
spiral_dataset_test();
load_trained_spiral_model_and_test();
let duration = start.elapsed();
println!("The benchmark took {:?}", duration);
}
pub fn spiral_dataset_test() {
// println!("generating spiral dataset");
//
// let (data, labels) = generate_spiral_dataset(3000, 3);
//
// let layer1 = Layer::init(2, 30, true);
// let layer2 = Layer::init(30, 30, true);
// let layer3 = Layer::init(30, 3, false);
//
// let layers = vec![layer1, layer2, layer3];
// let adam = Optimizer::Adam {
// learning_step: 0.05,
// beta1: 0.9,
// beta2: 0.999,
// };
//
// let mut model = Model::init(layers, adam, 0.001);
// model.train(&data, &labels, 50, 2, 500, 10, false, false);
// save_load::save_model(&model, "spiral_model".to_string()).unwrap();
}
pub fn load_trained_spiral_model_and_test() {
// let (test_data, test_labels) = generate_spiral_dataset(3000, 3);
// let mut spiral_model = save_load::load_model("spiral_model".to_string()).unwrap();
//
// let acc_trained: f64 = spiral_model.accuracy(&test_data, &test_labels);
//
// let layer1 = Layer::init(2, 30, true);
// let layer2 = Layer::init(30, 30, true);
// let layer3 = Layer::init(30, 3, false);
//
// let layers = vec![layer1, layer2, layer3];
// let adam = Optimizer::Adam {
// learning_step: 0.05,
// beta1: 0.9,
// beta2: 0.999,
// };
//
// let mut model = Model::init(layers, adam, 0.001);
// let acc_not_trained: f64 = model.accuracy(&test_data, &test_labels);
//
// println!("acc trained : {}", acc_trained);
// println!("acc not trained : {}", acc_not_trained);
}