PyTorch深度学习实践——9.多分类问题

课程链接：《PyTorch深度学习实践》9.多分类问题

课上示例

示例程序源代码+注释（根据个人理解）

import torch
from torchvision import transforms      # 图像处理工具
from torchvision import datasets
from torch.utils.data import DataLoader
import torch.nn.functional as F         # 使用 ReLU 作为激活函数
import torch.optim as optim             # 使用优化器

# prepare dataset
batch_size = 64
# 将单通道的像素值[0, 255]转变成像素张量[0, 1]，归一化；      均值，标准差
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)

# design model using class
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)  # 改变维数，变为二阶张量（矩阵）；-1其实就是自动获取mini_batch
        x = F.relu(self.l1(x))
        x = F.relu(self.l2(x))
        x = F.relu(self.l3(x))
        x = F.relu(self.l4(x))
        return self.l5(x)  # 最后一层不做激活，不进行非线性变换，直接传入softmax

model = Net()

# construct loss and optimizer
criterion = torch.nn.CrossEntropyLoss()
optimizer = optim.SGD(model.parameters(), lr=0.01, momentum=0.5)


# training cycle forward, backward, update
def train(epoch):               # 每轮循环封装成函数
    running_loss = 0.0
    for batch_idx, data in enumerate(train_loader, 0):
        inputs, target = data       # 数据x, y
        optimizer.zero_grad()       # 梯度清零

        outputs = model(inputs)             # 前馈
        loss = criterion(outputs, target)   # 计算损失
        loss.backward()                     # 反馈
        optimizer.step()                    # 优化

        running_loss += loss.item()         # 累计loss，使用.item()取loss
        if batch_idx % 300 == 299:          # 每300轮输出一次
            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():               # 不需要计算梯度
        for data in test_loader:
            images, labels = data
            outputs = model(images)

            # 沿第一个维度找最大值，返回最大值及其下标，下标用作标签
            _, predicted = torch.max(outputs.data, dim=1)  # dim = 1 列是第0个维度，行是第1个维度

            total += labels.size(0)
            correct += (predicted == labels).sum().item()  # 张量之间的比较运算，预测对的数量
    print('accuracy on test set: %d %% ' % (100 * correct / total))         # 正确数 correct ÷ 总数 total = 准确率 accuracy


if __name__ == '__main__':
    for epoch in range(10):     # 循环训练10个Epoch
        train(epoch)
        test()

课后练习

阅读文档，学习NLLLoss与CrossEntropyLoss区别

CrossEntropyLoss就是把以上Softmax–Log–NLLLoss合并成一步，参考链接。

课后作业

Otto Group Product Classification Challenge