用pytorch写一个DeepLab v3+网络完整代码，并用金字塔卷积模块替换其中的DCNN模块，最后解释如何使用这个网络

以下是用PyTorch写的一个DeepLab v3 网络完整代码，使用金字塔卷积模块替换了原本的DCNN模块：

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

class PyramidPoolingModule(nn.Module):
    def __init__(self, in_channels, pool_sizes):
        super(PyramidPoolingModule, self).__init__()
        self.pool_sizes = pool_sizes
        self.avg_pools = nn.ModuleList()
        for size in pool_sizes:
            self.avg_pools.append(nn.AdaptiveAvgPool2d(size))

    def forward(self, x):
        pools = [pool(x) for pool in self.avg_pools]
        pools = [F.interpolate(pool, size=(x.size(-2), x.size(-1)), mode='bilinear', align_corners=True) for pool in pools]
        out = torch.cat(pools, dim=1)
        return out

class DeepLabv3(nn.Module):
    def __init__(self, num_classes=21):
        super(DeepLabv3, self).__init__()
        self.backbone = nn.Sequential(
            nn.Conv2d(3, 64, 3, padding=1),
            nn.BatchNorm2d(64),
            nn.ReLU(),
            nn.Conv2d(64, 64, 3, padding=1),
            nn.BatchNorm2d(64),
            nn.ReLU(),
            nn.MaxPool2d(2, 2),

            nn.Conv2d(64, 128, 3, padding=1),
            nn.BatchNorm2d(128),
            nn.ReLU(),
            nn.Conv2d(128, 128, 3, padding=1),
            nn.BatchNorm2d(128),
            nn.ReLU(),
            nn.MaxPool2d(2, 2),

            nn.Conv2d(128, 256, 3, padding=1),
            nn.BatchNorm2d(256),
            nn.ReLU(),
            nn.Conv2d(256, 256, 3, padding=1),
            nn.BatchNorm2d(256),
            nn.ReLU(),
            nn.Conv2d(256, 256, 3, padding=1),
            nn.BatchNorm2d(256),
            nn.ReLU(),
            nn.MaxPool2d(2, 2),

            nn.Conv2d(256, 512, 3, padding=1),
            nn.BatchNorm2d(512),
            nn.ReLU(),
            nn.Conv2d(512, 512, 3, padding=1),
            nn.BatchNorm2d(512),