pytorch学习－－卷积神经网络

代码源自：https://github.com/yunjey/pytorch-tutorial
这里只是将其做为一个学习样例．

＃引入包
import torch 
import torch.nn as nn
import torchvision
import torchvision.transforms as transforms

                
# Device configuration　有显卡默认加载cuda0，否则用cpu
device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')

# Hyper parameters
num_epochs = 5
num_classes = 10 #分类数目
batch_size = 100 
learning_rate = 0.001#学习率

# MNIST dataset
train_dataset = torchvision.datasets.MNIST(root='../../data/',
                                           train=True, 
                                           transform=transforms.ToTensor(),
                                           download=True)

test_dataset = torchvision.datasets.MNIST(root='../../data/',
                                          train=False, 
                                          transform=transforms.ToTensor())

# Data loader
train_loader = torch.utils.data.DataLoader(dataset=train_dataset,
                                           batch_size=batch_size, 
                                           shuffle=True)

test_loader = torch.utils.data.DataLoader(dataset=test_dataset,
                                          batch_size=batch_size, 
                                          shuffle=False)

# Convolutional neural network (two convolutional layers)
#各个层的输入输出计算
#conv层　out = {(input_height - kernel_size + 2*padding) / stride[0] }+1 
#pooling   out ={ (input_height - kernel_size)/stride[0]} + 1

#Sequential就是个容器，把各个层放在一起组成一个大层
＃BatchNorm2d　BN层对于每个隐层神经元，把逐渐向非线性函数映射后向取值区间极限饱和区靠拢的输入分布强制
＃拉回到均值为0方差为1的比较标准的正态分布，使得非线性变换函数的输入值落入对输入比较敏感的区域，以此避免
＃梯度消失问题。
＃ReLU　激活层，非线性
＃MaxPool2d　层，　1. invariance(不变性)，这种不变性包括translation(平移)，rotation(旋转)，scale(尺度)
＃　2. 保留主要的特征同时减少参数(降维，效果类似PCA)和计算量，防止过拟合，提高模型泛化能力
class ConvNet(nn.Module):
    def __init__(self, num_classes=10):
        super(ConvNet, self).__init__()
        self.layer1 = nn.Sequential(
            nn.Conv2d(1, 16, kernel_size=5, stride=1, padding=2), # (28 - 5 + 2*2)/1 + 1 =  28
            nn.BatchNorm2d(16),
            nn.ReLU(),
            nn.MaxPool2d(kernel_size=2, stride=2)) #(28 - 2)/2 + 1 = 14
        self.layer2 = nn.Sequential(
            nn.Conv2d(16, 32, kernel_size=5, stride=1, padding=2),#(14 - 5 +4)/1 + 1 = 14
            nn.BatchNorm2d(32),
            nn.ReLU(),
            nn.MaxPool2d(kernel_size=2, stride=2)) #(14 - 2)/2 + 1 = 7
        self.fc = nn.Linear(7*7*32, num_classes)
        
    def forward(self, x):
        out = self.layer1(x)
        out = self.layer2(out)
        out = out.reshape(out.size(), -1)
        out = self.fc(out)
        return out

model = ConvNet(num_classes).to(device)

# Loss and optimizer 损失与优化器定义
criterion = nn.CrossEntropyLoss()
optimizer = torch.optim.Adam(model.parameters(), lr=learning_rate)

# Train the model
total_step = len(train_loader)＃每个epoch训练的次数
for epoch in range(num_epochs):
    for i, (images, labels) in enumerate(train_loader):
        images = images.to(device)
        labels = labels.to(device)
        
        # Forward pass　前向传播
        outputs = model(images)
        loss = criterion(outputs, labels)
        
        # Backward and optimize　反向传播及优化
        optimizer.zero_grad()#优化器每一次要先置０
        loss.backward()
        optimizer.step()
        
        if (i+1) % 100 == 0:
            print ('Epoch [{}/{}], Step [{}/{}], Loss: {:.4f}' 
                   .format(epoch+1, num_epochs, i+1, total_step, loss.item()))

# Test the model
model.eval()  # eval mode (batchnorm uses moving mean/variance instead of mini-batch mean/variance)
with torch.no_grad():
    correct = 0
    total = 0
    for images, labels in test_loader:
        images = images.to(device)
        labels = labels.to(device)
        outputs = model(images)
        _, predicted = torch.max(outputs.data, 1)＃例子中是(100,10)#这种输出结果， torch.max两个输出，最大值与
                                                                         #index,而我们只关心索引，做为结果输出，每一行的最大值
        total += labels.size(0)
        correct += (predicted == labels).sum().item()

    print('Test Accuracy of the model on the 10000 test images: {} %'.format(100 * correct / total))

# Save the model checkpoint
torch.save(model.state_dict(), 'model.ckpt')

基本流程：

１．下载并加载数据
２．定义网络
３．定义loss及优化器
４．定义训练各种参数
５．训练
６．测试集上测试
７．保存模型

[参考文档]
https://www.cnblogs.com/guoyaohua/p/8724433.html
https://blog.csdn.net/zxyhhjs2017/article/details/78607469