我有一个来自十个类别各100个共1000个的信号数据，每个数据有512个特征点，存储为一个(1000,1,512)的torch.tensor张量，现在我想将其输入一个深度DenseNet网络训练分类模型用于分类这些类别，请使用pytorch实现

以下是实现代码：

import torch
import torch.nn as nn
from torchvision.models import densenet121

# 定义DenseNet模型
class DenseNet(nn.Module):
    def __init__(self, num_classes):
        super(DenseNet, self).__init__()
        self.features = densenet121(pretrained=False).features
        self.classifier = nn.Linear(1024, num_classes)

    def forward(self, x):
        features = self.features(x)
        out = nn.functional.relu(features, inplace=True)
        out = nn.functional.adaptive_avg_pool2d(out, (1, 1)).view(features.size(0), -1)
        out = self.classifier(out)
        return out

# 数据准备
data = torch.randn(1000, 1, 512)
labels = torch.randint(0, 10, (1000,))

# 模型定义及训练
model = DenseNet(num_classes=10)
criterion = nn.CrossEntropyLoss()
optimizer = torch.optim.Adam(model.parameters())

for epoch in range(10):
    running_loss = 0.0
    for i in range(0, len(data), 64):
        inputs = data[i:i+64]
        targets = labels[i:i+64]

        optimizer.zero_grad()

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

        running_loss += loss.item()
    print(f"Epoch {epoch+1}, Loss: {running_loss/len(data):.4f}")

这里使用了预训练的DenseNet121作为特征提取器，然后添加一个全连接层进行分类。在训练过程中，使用交叉熵损失函数和Adam优化器进行模型优化。