학습 목표
- 다중 퍼셉트론(Multi Layer Perceptron)
- 오차역전파(Backpropagation)
1
import torch
1
2
3
4
5
6
device = 'cuda' if torch.cuda.is_available() else 'cpu'
# seed 고정
torch.manual_seed(777)
if device == 'cuda':
torch.cuda.manual_seed_all(777)
1
2
X = torch.FloatTensor([[0, 0], [0, 1], [1, 0], [1, 1]]).to(device)
Y = torch.FloatTensor([[0], [1], [1], [0]]).to(device)
nn Layers
두 개의 레이어를 가지는 다중 퍼셉트론
1
2
3
4
# nn Layers
linear1 = torch.nn.Linear(2, 2, bias=True)
linear2 = torch.nn.Linear(2, 1, bias=True)
sigmoid = torch.nn.Sigmoid()
1
model = torch.nn.Sequential(linear1, sigmoid, linear2, sigmoid).to(device)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
# Binary Cross-Entropy Loss
criterion = torch.nn.BCELoss().to(device)
optimizer = torch.optim.SGD(model.parameters(), lr=1)
for step in range(10001):
optimizer.zero_grad()
hypothesis = model(X)
cost = criterion(hypothesis, Y)
cost.backward()
optimizer.step()
# 100번에 한 번씩 로그 출력
if step % 100 == 0:
print(step, cost.item())
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
91
92
93
94
95
96
97
98
99
100
101
0 0.7434073090553284
100 0.693165123462677
200 0.6931577920913696
300 0.6931517124176025
400 0.6931463479995728
500 0.6931411027908325
600 0.6931357383728027
700 0.6931294798851013
800 0.6931220889091492
900 0.6931126117706299
1000 0.6930999755859375
1100 0.693082332611084
1200 0.6930569410324097
1300 0.6930190324783325
1400 0.6929606199264526
1500 0.6928660273551941
1600 0.6927032470703125
1700 0.6923960447311401
1800 0.6917302012443542
1900 0.6899653673171997
2000 0.6838315725326538
2100 0.6561667919158936
2200 0.4311023950576782
2300 0.1348935216665268
2400 0.06630441546440125
2500 0.042168211191892624
2600 0.030453883111476898
2700 0.023665912449359894
2800 0.01927776448428631
2900 0.01622403785586357
3000 0.01398380845785141
3100 0.012273931875824928
3200 0.010928118601441383
3300 0.009842472150921822
3400 0.008949032984673977
3500 0.008201336488127708
3600 0.007566767744719982
3700 0.007021686062216759
3800 0.006548595614731312
3900 0.006134224124252796
4000 0.005768375005573034
4100 0.0054430365562438965
4200 0.005151905119419098
4300 0.004889918025583029
4400 0.0046528722159564495
4500 0.004437457304447889
4600 0.004240859299898148
4700 0.004060701932758093
4800 0.003895031288266182
4900 0.0037421947345137596
5000 0.0036007347516715527
5100 0.003469479735940695
5200 0.0033473046496510506
5300 0.0032333978451788425
5400 0.0031268750317394733
5500 0.0030270610004663467
5600 0.0029333550482988358
5700 0.0028452035039663315
5800 0.00276215560734272
5900 0.002683777129277587
6000 0.0026096487417817116
6100 0.0025394847616553307
6200 0.0024729417636990547
6300 0.0024097643326967955
6400 0.002349698217585683
6500 0.0022925634402781725
6600 0.002238075714558363
6700 0.002186085097491741
6800 0.0021364721469581127
6900 0.002089011948555708
7000 0.00204361486248672
7100 0.0020001311786472797
7200 0.0019584265537559986
7300 0.0019184107659384608
7400 0.0018799942918121815
7500 0.0018430722411721945
7600 0.0018075549742206931
7700 0.0017733527347445488
7800 0.0017404207028448582
7900 0.0017087137093767524
8000 0.001678097527474165
8100 0.0016485571395605803
8200 0.0016200175741687417
8300 0.0015924491453915834
8400 0.0015657918993383646
8500 0.0015400308184325695
8600 0.0015150614781305194
8700 0.0014909137971699238
8800 0.0014674977865070105
8900 0.001444813678972423
9000 0.0014228165382519364
9100 0.0014014765620231628
9200 0.0013806892093271017
9300 0.0013606036081910133
9400 0.0013410557294264436
9500 0.001322030322626233
9600 0.001303557539358735
9700 0.001285637030377984
9800 0.0012681199004873633
9900 0.0012511102249845862
10000 0.0012345188297331333
1
2
3
4
5
6
7
with torch.no_grad():
hypothesis = model(X)
predicted = (hypothesis > 0.5).float()
accuracy = (predicted == Y).float().mean()
print('\nHypothesis: ', hypothesis.detach().cpu().numpy(),
'\nCorrect: ', predicted.detach().cpu().numpy(),
'\nAccuracy: ', accuracy.item())
1
2
3
4
5
6
7
8
9
Hypothesis: [[0.00106364]
[0.99889404]
[0.99889404]
[0.00165862]]
Correct: [[0.]
[1.]
[1.]
[0.]]
Accuracy: 1.0
nn Wide Deep Layers
네 개의 레이어를 가지는 다중 퍼셉트론
1
2
3
4
5
6
# nn wide deep Layers
linear1 = torch.nn.Linear(2, 10, bias=True)
linear2 = torch.nn.Linear(10, 10, bias=True)
linear3 = torch.nn.Linear(10, 10, bias=True)
linear4 = torch.nn.Linear(10, 1, bias=True)
sigmoid = torch.nn.Sigmoid()
1
model = torch.nn.Sequential(linear1, sigmoid, linear2, sigmoid, linear3, sigmoid, linear4, sigmoid).to(device)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
# Binary Cross-Entropy Loss
criterion = torch.nn.BCELoss().to(device)
optimizer = torch.optim.SGD(model.parameters(), lr=1)
for step in range(10001):
optimizer.zero_grad()
hypothesis = model(X)
cost = criterion(hypothesis, Y)
cost.backward()
optimizer.step()
# 1000번에 한 번씩 로그 출력
if step % 1000 == 0:
print(step, cost.item())
1
2
3
4
5
6
7
8
9
10
11
0 0.6978083848953247
1000 0.6929439902305603
2000 0.6797102689743042
3000 0.002333962358534336
4000 0.0006081871688365936
5000 0.00033300856011919677
6000 0.0002254210994578898
7000 0.0001689193450147286
8000 0.00013438795576803386
9000 0.00011119883856736124
10000 9.461208537686616e-05
1
2
3
4
5
6
7
with torch.no_grad():
hypothesis = model(X)
predicted = (hypothesis > 0.5).float()
accuracy = (predicted == Y).float().mean()
print('\nHypothesis: ', hypothesis.detach().cpu().numpy(),
'\nCorrect: ', predicted.detach().cpu().numpy(),
'\nAccuracy: ', accuracy.item())
1
2
3
4
5
6
7
8
9
Hypothesis: [[7.4425239e-05]
[9.9988043e-01]
[9.9990392e-01]
[8.8329281e-05]]
Correct: [[0.]
[1.]
[1.]
[0.]]
Accuracy: 1.0