《PyTorch深度学习实践》（4）

主题

多分类问题

总结

多分类问题各个分量的输出应该是相互抑制的，带有竞争性的。
softmax层中同层的节点输入会影响当前节点的输出。
CrossEntropyLoss = LogSoftmax + NLLLOSS。采用CrossEntropyLoss中包含有LogSoftmax激活层，所以模型的最后一层不设置激活

NLLLOSS.png

CrossEntropyLoss.png

transform = transforms.Compose([··· , ···])：建立dataset时传入的参数，用于图像变换。python读取图像的格式是PIL
图像文件中像素存储结构是WHC，需要转换成CWH
transforms.Normalize(( , ) , ( , ))：把数据变换至标准正态分布
x = x.view(-1, 28*28)：数据尺寸转换节点，把图片平铺使得可以作为nn.Linear的输入

代码

接下来介绍完整代码，分为：

导入模块
主代码
数据准备
定义模型
创建模型、损失函数节点、优化器
定义训练/测试过程
执行结果

1. 导入模块

import torchvision.transforms as transforms
import torchvision.datasets as datasets
from torch.utils.data import DataLoader
import torch.nn.functional as F
import torch

2. 主代码

if __name__ == '__main__':
    for epoch in range(10):
        train(epoch)
        test()

3. 数据准备

batch_size = 64
transform = transforms.Compose([transforms.ToTensor(),
                                transforms.Normalize((0.1307,), (0.3081,))])

train_dataset = datasets.MNIST(root='../dataset/mnist/',
                               train=True,
                               download=True,
                               transform=transform)
train_loader = DataLoader(train_dataset,
                          shuffle=True,
                          batch_size=batch_size)

test_dataset = datasets.MNIST(root='../dataset/mnist/',
                              train=False,
                              download=True,
                              transform=transform)
test_loader = DataLoader(test_dataset,
                         shuffle=False,
                         batch_size=batch_size)

4. 定义模型

class Net(torch.nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.l1 = torch.nn.Linear(784, 512)
        self.l2 = torch.nn.Linear(512, 256)
        self.l3 = torch.nn.Linear(256, 128)
        self.l4 = torch.nn.Linear(128, 64)
        self.l5 = torch.nn.Linear(64, 10)

    def forward(self, x):
        x = x.view(-1, 784)
        x = F.relu(self.l1(x))
        x = F.relu(self.l2(x))
        x = F.relu(self.l3(x))
        x = F.relu(self.l4(x))
        x = self.l5(x)
        return x

5. 创建模型、损失函数节点、优化器

model = Net()
criterion = torch.nn.CrossEntropyLoss()
optimizer = torch.optim.SGD(model.parameters(), lr=0.01, momentum=0.5)

6. 定义训练/测试过程

def train(epoch):
    running_loss = 0.0
    for batch_idx, data in enumerate(train_loader, 0):
        inputs, target = data
        optimizer.zero_grad()

        outputs = model(inputs)
        loss = criterion(outputs, target)
        loss.backward()
        optimizer.step()

        running_loss += loss.item()
        if batch_idx % 300 == 299:
            print('[%d, %5d] Loss: %.3f' % (epoch+1, batch_idx+1, running_loss/300))
            running_loss = 0.0

def test():
    correct = 0
    total = 0
    with torch.no_grad():
        for data in test_loader:
            images, labels = data
            outputs = model(images)
            _, predict = torch.max(outputs.data, dim=1)
            total += labels.size(0)
            correct += (predict == labels).sum().item()
    print('Accuracy on test set:%d %%' % (100 * correct/total))

7. 执行结果

[1,   300] Loss: 2.227
[1,   600] Loss: 1.038
[1,   900] Loss: 0.436
Accuracy on test set:89 %
[2,   300] Loss: 0.329
[2,   600] Loss: 0.273
[2,   900] Loss: 0.247
Accuracy on test set:94 %
[3,   300] Loss: 0.197
[3,   600] Loss: 0.175
[3,   900] Loss: 0.158
Accuracy on test set:95 %
[4,   300] Loss: 0.135
[4,   600] Loss: 0.125
[4,   900] Loss: 0.116
Accuracy on test set:96 %
[5,   300] Loss: 0.095
[5,   600] Loss: 0.098
[5,   900] Loss: 0.094
Accuracy on test set:96 %
[6,   300] Loss: 0.074
[6,   600] Loss: 0.078
[6,   900] Loss: 0.075
Accuracy on test set:97 %
[7,   300] Loss: 0.062
[7,   600] Loss: 0.060
[7,   900] Loss: 0.058
Accuracy on test set:96 %
[8,   300] Loss: 0.048
[8,   600] Loss: 0.052
[8,   900] Loss: 0.048
Accuracy on test set:97 %
[9,   300] Loss: 0.037
[9,   600] Loss: 0.042
[9,   900] Loss: 0.041
Accuracy on test set:97 %
[10,   300] Loss: 0.032
[10,   600] Loss: 0.035
[10,   900] Loss: 0.032
Accuracy on test set:97 %

Process finished with exit code 0