-
Notifications
You must be signed in to change notification settings - Fork 4
/
Copy pathLFlatten.h
90 lines (53 loc) · 1.47 KB
/
LFlatten.h
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
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
#ifndef FLATTENLAYER_H
#define FLATTENLAYER_H
#include "Layer.h"
class LFlatten : public Layer {
private:
//Gradients
Tensor dCdX;
public:
//Constructor
LFlatten( int in_dim, int in_rows, int in_cols ) {
//Set dimensions
this->in_dim = in_dim;
this->in_rows = in_rows;
this->in_cols = in_cols;
this->out_dim = 1;
out_rows = in_dim * in_rows * in_cols;
out_cols = 1;
//Redimension matrices
in.resize(in_dim, in_rows, in_cols);
out.resize(out_dim, out_rows, out_cols);
dCdX.resize(in_dim, in_rows, in_cols);
}
//Properties
char getType() { return 'f'; }
//Functions
Tensor feedforward( Tensor in ) {
this->in = in.copy();
int count = 0;
for (int d = 0; d < in_dim; d++) {
for (int i = 0; i < in_rows; i++) {
for (int j = 0; j < in_cols; j++) {
out(0, count, 0) = in(d, i, j);
count++;
}
}
}
return out;
}
Tensor feedback( Tensor delta ) {
int count = 0;
for (int d = 0; d < in_dim; d++) {
for (int i = 0; i < in_rows; i++) {
for (int j = 0; j < in_cols; j++) {
dCdX(d, i, j) = delta(0, count, 0);
count++;
}
}
}
return dCdX;
}
void updateweights( float rate, float mom ) { return; }
};
#endif