PyTorch笔记 入门：写一个简单的神经网络1：Regression（回归）_pytorch regressionlayer

注意：本文无理论和推导
相关视频：
[双语字幕]吴恩达深度学习deeplearning.ai
PyTorch 动态神经网络 (莫烦 Python 教学)

一、导入库、创造数据

二、基本格式

三、编写网络

1.设计网络

实现左边那个三层的简单网络
input layer的神经元个数：n_input
hidden layer的神经元个数：n_hidden
output layer的神经元个数：n_output

有些人又将n_input写为n_features

2.init（初始化）

3.forward

4.创建网络

5.优化器

6.训练网络

7.画出loss的图像

8.预测

四、完整代码

import torch
import torch.nn.functional as F
from torch.autograd import Variable
import matplotlib.pyplot as plt

# 自己创建的网路
class Net(torch.nn.Module):
    def __init__(self, n_input, n_hidden, n_output):
        super(Net, self).__init__()
        self.hidden_layer = torch.nn.Linear(n_input, n_hidden)
        self.output_layer = torch.nn.Linear(n_hidden, n_output)
    
    def forward(self, x):
        x = F.relu(self.hidden_layer(x))
        x = self.output_layer(x)
        return x

# 创造数据
x = torch.unsqueeze(torch.linspace(-1,1,100), dim=1)
print(x.shape)

y = torch.pow(x, 2) + 0.2 * torch.rand(x.size())
print(y.shape)

plt.scatter(x.data.numpy(), y.data.numpy())
plt.show()

x, y = Variable(x), Variable(y)

# 创建网络
net = Net(1, 10, 1)
print(net)

# 优化器
optimizer = torch.optim.SGD(net.parameters(), lr=0.5)
# 损失函数
loss_func = torch.nn.MSELoss() # Mean Square Error Loss

# 训练
losses = []
for i in range(100):
    # 计算loss
    predict = net(x)
    loss = loss_func(predict, y)
    # 训练优化
    optimizer.zero_grad()
    loss.backward()
    optimizer.step()
    # 每训练5次输出一次loss
    if i%5 == 0:
        print(loss.data.item())
        losses.append(loss.data.item())
        
# 画出loss的图像
plt.plot(range(len(losses)), losses)
plt.xlabel('epoch')
plt.ylabel('loss')
plt.xticks(range(0,20,2))
plt.show()

# 预测
test = torch.tensor([[1.1], [1.2], [1.3], [1.4], [1.5]])
print(test)

result = net(test)
print(result)

# 实际与预测的差距
real = torch.pow(test, 2)
print(real)

print(real - result)
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