densenet，测试代码pytorch实现

DenseNet，全称为密集连接网络（Densely Connected Convolutional Networks），是由Gao Huang等人于2016年提出的深度学习模型。它的一个显著特点是前一层的输出作为输入直接馈送到下一层，这样每个层都能利用到前面所有层的信息，增加了信息流动和特征复用，有助于解决梯度消失问题并加速训练。

在PyTorch中实现DenseNet的基本步骤包括：

1. 导入必要的库：

import torch
import torch.nn as nn
from torch.nn import functional as F

2. 定义基础块（Basic Block）：

class BasicBlock(nn.Module):
    def __init__(self, in_channels, growth_rate):
        super(BasicBlock, self).__init__()
        self.conv1 = nn.Conv2d(in_channels, growth_rate, kernel_size=3, padding=1)
        self.bn1 = nn.BatchNorm2d(growth_rate)
        self.relu = nn.ReLU(inplace=True)
        self.conv2 = nn.Conv2d(growth_rate, growth_rate, kernel_size=3, padding=1)
        self.bn2 = nn.BatchNorm2d(growth_rate)

    def forward(self, x):
        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)
        out = self.conv2(out)
        out = self.bn2(out)
        return torch.cat([x, out], 1)  # Concatenate along the channel dimension

3. 实现DenseNet结构：

class DenseNet(nn.Module):
    def __init__(self, num_blocks, growth_rate, num_classes):
        super(DenseNet, self).__init__()
        self.growth_rate = growth_rate
        self.num_classes = num_classes
        self.features = self.create_dense_layers(num_blocks, num_classes)
        self.classifier = nn.Linear(num_features * growth_rate, num_classes)

    def create_dense_layers(self, num_blocks, num_classes):
        layers = []
        in_channels = 2 * growth_rate  # Initial channels for first block
        for _ in range(num_blocks):
            layers.append(BasicBlock(in_channels, growth_rate))
            in_channels += growth_rate
        layers.append(nn.Conv2d(in_channels, num_classes, kernel_size=1))  # Output layer
        return nn.Sequential(*layers)

    def forward(self, x):
        features = self.features(x)
        out = F.avg_pool2d(features, features.size()[2:])
        out = out.view(out.size(0), -1)
        out = self.classifier(out)
        return out