【动手学习pytorch笔记】5.处理过拟合之——Dropout

Dropout

定义Dropout函数

import torch
from torch import nn
from d2l import torch as d2l


def dropout_layer(X, dropout):
    assert 0 <= dropout <= 1

    # 在本情况中，所有元素都被丢弃

    if dropout == 1:
        return torch.zeros_like(X)

    # 在本情况中，所有元素都被保留

    if dropout == 0:
        return X
    mask = (torch.rand(X.shape) > dropout).float()
    return mask * X / (1.0 - dropout)

mask用来生成 0 或 1 的随机数。

mask * X / (1.0 - dropout) 用来保证期望不发生变化。

测试一下

X= torch.arange(16, dtype = torch.float32).reshape((2, 8))
print(X)
print(dropout_layer(X, 0.))
print(dropout_layer(X, 0.5))
print(dropout_layer(X, 1.))

tensor([[ 0.,  1.,  2.,  3.,  4.,  5.,  6.,  7.],
        [ 8.,  9., 10., 11., 12., 13., 14., 15.]])
tensor([[ 0.,  1.,  2.,  3.,  4.,  5.,  6.,  7.],
        [ 8.,  9., 10., 11., 12., 13., 14., 15.]])
tensor([[ 0.,  0.,  4.,  6.,  0.,  0.,  0.,  0.],
        [16., 18., 20.,  0.,  0., 26.,  0., 30.]])
tensor([[0., 0., 0., 0., 0., 0., 0., 0.],
        [0., 0., 0., 0., 0., 0., 0., 0.]])

从零开始实现

定义模型参数

num_inputs, num_hiddens1, num_hiddens2, num_outputs = 784, 256, 256，10

两个隐藏层的感知机，即有三个线性层

定义模型

dropout1, dropout2 = 0.2, 0.5

class Net(nn.Module):
    def __init__(self, num_inputs, num_outputs, num_hiddens1, num_hiddens2,
                 is_training = True):
        super(Net, self).__init__()
        self.num_inputs = num_inputs
        self.training = is_training
        self.lin1 = nn.Linear(num_inputs, num_hiddens1)
        self.lin2 = nn.Linear(num_hiddens1, num_hiddens2)
        self.lin3 = nn.Linear(num_hiddens2, num_outputs)
        self.relu = nn.ReLU()

    def forward(self, X):
        H1 = self.relu(self.lin1(X.reshape((-1, self.num_inputs))))
        # 只有在训练模型时才使用dropout
        if self.training == True:
            # 在第一个全连接层之后添加一个dropout层
            H1 = dropout_layer(H1, dropout1)
        H2 = self.relu(self.lin2(H1))
        if self.training == True:
            # 在第二个全连接层之后添加一个dropout层
            H2 = dropout_layer(H2, dropout2)
        out = self.lin3(H2)
        return out


net = Net(num_inputs, num_outputs, num_hiddens1, num_hiddens2)

训练

num_epochs, lr, batch_size = 10, 0.5, 256
loss = nn.CrossEntropyLoss(reduction='none')
train_iter, test_iter = d2l.load_data_fashion_mnist(batch_size)
trainer = torch.optim.SGD(net.parameters(), lr=lr)
d2l.train_ch3(net, train_iter, test_iter, loss, num_epochs, trainer)

​

简易实现

定义模型

net = nn.Sequential(nn.Flatten(),
        			nn.Linear(784, 256),
        			nn.ReLU(),
       				# 在第一个全连接层之后添加一个dropout层
        			nn.Dropout(dropout1),
        			nn.Linear(256, 256),
        			nn.ReLU(),
        			# 在第二个全连接层之后添加一个dropout层
        			nn.Dropout(dropout2),
        			nn.Linear(256, 10))

#初始化权重
def init_weights(m):
    if type(m) == nn.Linear:
        nn.init.normal_(m.weight, std=0.01)

net.apply(init_weights);

输出层并不需要Dropout

训练

trainer = torch.optim.SGD(net.parameters(), lr=lr)
d2l.train_ch3(net, train_iter, test_iter, loss, num_epochs, trainer)