深度学习算法之cifar10数据集训练和测试步骤以及相关代码Pytorch

今天忽然看到 CIFAR10的分类，自己就直接进行了相关备注，给初学者分享一下，如果代码跑不通，可以评论区反馈。一起进步。

整体步骤就是下面5部分。

1. 使用torchvision加载并标准化 CIFAR10 训练和测试数据集
2. 定义卷积神经网络
3. 定义损失函数
4. 根据训练数据训练网络
5. 在测试数据上测试网络

1、首先导入包

torchvision.transforms用于进行图片变换（归一化等操作），torchvision含有一些数据集可以直接下载使用，matplotlib用于画图的包，torch.optim用于优化器部分。

import torch
import torchvision
import torchvision.transforms as transforms
import matplotlib.pyplot as plt
import numpy as np
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim

二、加载数据集 并查看数据集

transform = transforms.Compose(
    [transforms.ToTensor(),#模型只接受tensor格式
     transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])#进行平均值，方差的标准化

trainset = torchvision.datasets.CIFAR10(root='./data', train=True,
                                        download=True, transform=transform)#如果为download=False则表示不需要下载，但是本地必须要有相应的文件
trainloader = torch.utils.data.DataLoader(trainset, batch_size=4,
                                          shuffle=True, num_workers=2)
#shuffle=True是将数据集的图片打乱，这样有利于训练
testset = torchvision.datasets.CIFAR10(root='./data', train=False,
                                       download=True, transform=transform)
testloader = torch.utils.data.DataLoader(testset, batch_size=4,
                                         shuffle=False, num_workers=2)

classes = ('plane', 'car', 'bird', 'cat',#这是该数据的类别
           'deer', 'dog', 'frog', 'horse', 'ship', 'truck')

#展示图片
def imshow(img):
    img = img / 2 + 0.5     # unnormalize 归一化操作
    npimg = img.numpy()
    plt.imshow(np.transpose(npimg, (1, 2, 0)))
    plt.show()

# get some random training images
dataiter = iter(trainloader)
images, labels = dataiter.next()#这是一个迭代器。

# show images
imshow(torchvision.utils.make_grid(images))
# print labels
print(' '.join('%5s' % classes[labels[j]] for j in range(4)))

三、定义模型 

#这个模型比较简单，大家可以更换复杂模型，只要维度对上就行
class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(3, 6, 5) #输入通道数3，输出通道数为6，卷积核为5x5大小
        self.pool = nn.MaxPool2d(2, 2)#最大池化层
        self.conv2 = nn.Conv2d(6, 16, 5)
        self.fc1 = nn.Linear(16 * 5 * 5, 120)#全连接层，必须将[16，5 ， 5]先view()成16 * 5 * 5才能使用全连接层
        self.fc2 = nn.Linear(120, 84)
        self.fc3 = nn.Linear(84, 10)

    def forward(self, x):
        x = self.pool(F.relu(self.conv1(x)))
        x = self.pool(F.relu(self.conv2(x)))
        x = x.view(-1, 16 * 5 * 5)
        x = F.relu(self.fc1(x))
        x = F.relu(self.fc2(x))
        x = self.fc3(x)
        return x

net = Net()#实例化网络模型
#损失函数为交叉熵损失函数
#采用的是优化器是随机梯度下降优化器（是包含动量部分的）
criterion = nn.CrossEntropyLoss()
optimizer = optim.SGD(net.parameters(), lr=0.001, momentum=0.9)

 四、进行训练

for epoch in range(2):  # 所需要训练的轮次

    running_loss = 0.0
    for i, data in enumerate(trainloader, 0):
        # get the inputs; data = [inputs, labels]
        inputs, labels = data
        #得到输入图片和标签
        # 梯度需要清零，否则会干扰模型
        optimizer.zero_grad()

        # forward + backward + optimize
        outputs = net(inputs)
        loss = criterion(outputs, labels)
        loss.backward()#反项传播
        optimizer.step()

        # print statistics
        running_loss += loss.item()#.item()这也属于迭代器，将每一步的损失取出来
        if i % 2000 == 1999:    # print every 2000 mini-batches
            print('[%d, %5d] loss: %.3f' %
                  (epoch + 1, i + 1, running_loss / 2000))
            running_loss = 0.0

print('Finished Training')

 五、保存模型并进行验证

PATH = './cifar_net.pth'#权重路径
torch.save(net.state_dict(), PATH)#保存所有权重，也有保存部分权重的，大家可以去网上找找

dataiter = iter(testloader)
images, labels = dataiter.next()

# print images
imshow(torchvision.utils.make_grid(images))
print('GroundTruth: ', ' '.join('%5s' % classes[labels[j]] for j in range(4)))

net = Net()
net.load_state_dict(torch.load(PATH))#加载训练好的模型
outputs = net(images)
_, predicted = torch.max(outputs, 1)

print('Predicted: ', ' '.join('%5s' % classes[predicted[j]]
                              for j in range(4)))
correct = 0
total = 0
with torch.no_grad():
    for data in testloader:
        images, labels = data
        outputs = net(images)
        _, predicted = torch.max(outputs.data, 1)
        total += labels.size(0)
        correct += (predicted == labels).sum().item()

print('Accuracy of the network on the 10000 test images: %d %%' % (
    100 * correct / total))

 用于测试每一个类别所对应的概率

class_correct = list(0\. for i in range(10))
class_total = list(0\. for i in range(10))
with torch.no_grad():
    for data in testloader:
        images, labels = data
        outputs = net(images)
        _, predicted = torch.max(outputs, 1)
        c = (predicted == labels).squeeze()
        for i in range(4):
            label = labels[i]
            class_correct[label] += c[i].item()
            class_total[label] += 1

for i in range(10):
    print('Accuracy of %5s : %2d %%' % (
        classes[i], 100 * class_correct[i] / class_total[i]))