pytorch工具——使用pytorch构建一个分类器

**PyTorch工具——使用PyTorch构建一个分类器**

在深度学习领域，PyTorch是一个非常流行的开源机器学习库。它提供了一个简单易用的接口，让开发者可以快速构建、训练和部署各种模型。以下是如何使用PyTorch构建一个基本的分类器。

###1. 安装PyTorch首先，我们需要安装PyTorch。如果你已经安装过PyTorch，可以跳过这一步。

bashpip install torch torchvision

import torchimport torch.nn as nnimport torch.optim as optimfrom torchvision import datasets, transforms

# 数据预处理transform = transforms.Compose([transforms.ToTensor()])

# 加载数据train_dataset = datasets.MNIST(root='./data', train=True, download=True, transform=transform)
test_dataset = datasets.MNIST(root='./data', train=False, download=True, transform=transform)

# 定义数据加载器batch_size =64train_loader = torch.utils.data.DataLoader(dataset=train_dataset, batch_size=batch_size, shuffle=True)
test_loader = torch.utils.data.DataLoader(dataset=test_dataset, batch_size=batch_size, shuffle=False)

class Net(nn.Module):
 def __init__(self):
 super(Net, self).__init__()
 # 卷积层 self.conv1 = nn.Conv2d(1,10, kernel_size=5)
 self.conv2 = nn.Conv2d(10,20, kernel_size=5)
 # 全连接层 self.fc1 = nn.Linear(320,50)
 self.fc2 = nn.Linear(50,10)

 def forward(self, x):
 # 卷积层 x = torch.relu(torch.max_pooling2d(self.conv1(x),2,2))
 x = torch.relu(torch.max_pooling2d(self.conv2(x),2,2))
 # 全连接层 x = x.view(-1,320)
 x = torch.relu(self.fc1(x))
 x = self.fc2(x)

 return x

# 损失函数criterion = nn.CrossEntropyLoss()

#优化器optimizer = optim.SGD(net.parameters(), lr=0.01, momentum=0.9)

for epoch in range(10): # 迭代10次 for i, data in enumerate(train_loader):
 inputs, labels = data # 前向传播 outputs = net(inputs)
 # 后向传播 loss = criterion(outputs, labels)
 # 更新参数 optimizer.zero_grad()
 loss.backward()
 optimizer.step()

 print('Epoch {}: Loss = {:.4f}'.format(epoch+1, loss.item()))

net.eval()
test_loss =0correct =0with torch.no_grad():
 for data in test_loader:
 inputs, labels = data outputs = net(inputs)
 loss = criterion(outputs, labels)
 test_loss += loss.item()
 _, predicted = torch.max(outputs,1)
 correct += (predicted == labels).sum().item()

accuracy = correct / len(test_dataset)
print('Test Loss: {:.4f}, Accuracy: {:.2f}%'.format(test_loss / len(test_loader), accuracy *100))