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
|
from .scalar import Scalar
import random
class Neuron:
def __init__(self, n_X):
self.n_X = n_X
self.w = [ Scalar(random.uniform(-1, 1)) for _ in range(n_X) ]
self.b = Scalar(random.uniform(-1, 1))
def __call__(self, X):
result = 0
for wi, Xi in zip(self.w, X):
result += wi * Xi
result += self.b
return result.tanh()
def parameters(self):
return self.w + [ self.b ]
class Layer:
def __init__(self, n_X, n_y):
self.neurons = [ Neuron(n_X) for _ in range(n_y) ]
def __call__(self, X):
result = [ n(X) for n in self.neurons ]
return result[0] if len(result) == 1 else result
def parameters(self):
return [ param for neuron in self.neurons for param in neuron.parameters() ]
class MLP:
def __init__(self, n_X, layers):
sz = [ n_X ] + layers
self.layers = [ Layer(sz[i], sz[i + 1]) for i in range(len(layers)) ]
def __call__(self, X):
for layer in self.layers:
X = layer(X)
return X
def parameters(self):
return [ param for layer in self.layers for param in layer.parameters() ]
def optimise(self, lr):
for parameter in self.parameters():
parameter.data -= lr * parameter.grad
|
