基于lenet的手写数字识别PYTORCH

LeNet是一种经典的卷积神经网络（Convolutional Neural Network，CNN），它最初由Yann LeCun等人在1998年应用于手写数字识别任务，尤其是MNIST数据集上。在PyTorch中实现基于LeNet的手写数字识别可以分为以下几个步骤：

1. **导入库**：首先，你需要导入必要的PyTorch库，包括`torch`, `torch.nn`, 和 `torch.optim`。

import torch
import torch.nn as nn
import torch.optim as optim

2. **构建模型**：定义一个继承自`nn.Module`的类，其中包含卷积层、池化层、全连接层等组件。

class LeNet(nn.Module):
    def __init__(self):
        super(LeNet, self).__init__()
        self.conv1 = nn.Conv2d(1, 6, kernel_size=5)
        self.pool = nn.MaxPool2d(2, 2)
        self.conv2 = nn.Conv2d(6, 16, kernel_size=5)
        self.fc1 = nn.Linear(16 * 4 * 4, 120)  # Flatten the output of conv2
        self.fc2 = nn.Linear(120, 84)
        self.fc3 = nn.Linear(84, 10)  # Output layer for 10 classes

    def forward(self, x):
        x = F.relu(self.pool(F.relu(self.conv1(x))))
        x = F.relu(self.pool(F.relu(self.conv2(x))))
        x = x.view(-1, 16 * 4 * 4)  # Reshape for fully connected layers
        x = F.relu(self.fc1(x))
        x = F.relu(self.fc2(x))
        x = self.fc3(x)
        return x

3. **加载数据**：使用PyTorch的数据集模块如`torchvision.datasets`获取MNIST数据，并进行预处理。

from torchvision import datasets, transforms
transform = transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.1307,), (0.3081,))])
train_dataset = datasets.MNIST(root='./data', train=True, transform=transform, download=True)
test_dataset = datasets.MNIST(root='./data', train=False, transform=transform)

train_loader = torch.utils.data.DataLoader(train_dataset, batch_size=64, shuffle=True)
test_loader = torch.utils.data.DataLoader(test_dataset, batch_size=64, shuffle=False)

4. **训练模型**：定义损失函数（如交叉熵）、优化器（如Adam）并开始训练过程。

model = LeNet()
criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(model.parameters(), lr=0.001)

for epoch in range(num_epochs):  # Train for a number of epochs
    for images, labels in train_loader:
        optimizer.zero_grad()  # Clear gradients
        outputs = model(images)
        loss = criterion(outputs, labels)
        loss.backward()  # Backpropagation
        optimizer.step()  # Update weights

5. **评估模型**：测试模型在测试集上的性能。

correct_count, all_count = 0, 0
with torch.no_grad():
    for images, labels in test_loader:
        outputs = model(images)
        _, predicted = torch.max(outputs.data, 1)
        all_count += labels.size(0)
        correct_count += (predicted == labels).sum().item()

accuracy = 100 * correct_count / all_count
print(f"Test accuracy: {accuracy}%")