利用pytorch，搭建一个十分类的卷积神经网络模型用子数字识别，数据集使用自己的，一次卷积必须有一次池化卷积，要求只使用一次Sequential函数将所有的层组合，并注释清楚，及import的库

好的，以下是代码实现：

import torch
import torch.nn as nn
import torch.optim as optim
import torchvision.transforms as transforms
from torch.data import DataLoader, Dataset
from PIL import Image
import os

# 定义数据集类
class MyDataset(Dataset):
    def __init__(self, data_dir):
        self.data_dir = data_dir
        self.transform = transforms.Compose([
            transforms.ToTensor(),
            transforms.Normalize(mean=[0.5], std=[0.5])
        ])
        self.imgs = os.listdir(data_dir)

    def __getitem__(self, index):
        img_path = os.path.join(self.data_dir, self.imgs[index])
        label = int(self.imgs[index].split("_")[0])
        img = Image.open(img_path)
        img = self.transform(img)
        return img, label

    def __len__(self):
        return len(self.imgs)

# 定义卷积神经网络模型
class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.layers = nn.Sequential(
            nn.Conv2d(1, 16, 3, padding=1), # 卷积层1
            nn.BatchNorm2d(16), # BatchNorm层1
            nn.ReLU(), # ReLU层1
            nn.MaxPool2d(2, 2), # 池化层1
            nn.Conv2d(16, 32, 3, padding=1), # 卷积层2
            nn.BatchNorm2d(32), # BatchNorm层2
            nn.ReLU(), # ReLU层2
            nn.MaxPool2d(2, 2), # 池化层2
            nn.Conv2d(32, 64, 3, padding=1), # 卷积层3
            nn.BatchNorm2d(64), # BatchNorm层3
            nn.ReLU(), # ReLU层3
            nn.MaxPool2d(2, 2), # 池化层3
            nn.Flatten(), # 拉平层
            nn.Linear(64 * 3 * 3, 128), # 全连接层1
            nn.ReLU(), # ReLU层4
            nn.Linear(128, 10) # 全连接层2
        )

    def forward(self, x):
        x = self.layers(x)
        return x

# 加载数据集
train_dataset = MyDataset("./train/") # 训练集
test_dataset = MyDataset("./test/") # 测试集
train_loader = DataLoader(train_dataset, batch_size=64, shuffle=True)
test_loader = DataLoader(test_dataset, batch_size=64, shuffle=True)

# 实例化模型并定义损失函数和优化器
model = Net()
criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(model.parameters(), lr=0.001)

# 训练模型
for epoch in range(10):
    running_loss = 0.0
    for i, data in enumerate(train_loader, 0):
        inputs, labels = data
        optimizer.zero_grad()
        outputs = model(inputs)
        loss = criterion(outputs, labels)
        loss.backward()
        optimizer.step()
        running_loss += loss.item()
        if i % 100 == 99:
            print('[%d, %5d] loss: %.3f' %
                  (epoch + 1, i + 1, running_loss / 100))
            running_loss = 0.0

print("Finished Training")

# 在测试集上评估模型
correct = 0
total = 0
with torch.no_grad():
    for data in test_loader:
        images, labels = data
        outputs = model(images)
        _, predicted = torch.max(outputs.data, 1)
        total += labels.size(0)
        correct += (predicted == labels).sum().item()

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

注释在代码中已经标明，import的库有：torch、torchvision、os、PIL。