帮我构建一个使用pytorch构建的mobilenetv2网络结构

下面是使用 PyTorch 构建的 MobileNetV2 的代码示例：

import torch
import torch.nn as nn

class MobileNetV2(nn.Module):
    def __init__(self, num_classes=1000):
        super(MobileNetV2, self).__init__()

        def make_divisible(v, divisor, min_value=None):
            if min_value is None:
                min_value = divisor
            new_v = max(min_value, int(v + divisor / 2) // divisor * divisor)
            if new_v < 0.9 * v:
                new_v += divisor
            return new_v

        # building the first layer
        input_channel = 32
        last_channel = make_divisible(input_channel * width_mult, 8)
        self.conv1 = nn.Conv2d(3, last_channel, kernel_size=3, stride=2, padding=1, bias=False)
        self.bn1 = nn.BatchNorm2d(last_channel)
        self.relu = nn.ReLU()

        # building the inverted residual blocks
        self.layer1 = self._make_layer(16, 1, last_channel)
        self.layer2 = self._make_layer(24, 6, last_channel)
        self.layer3 = self._make_layer(32, 6, last_channel)
        self.layer4 = self._make_layer(64, 6, last_channel)
        self.layer5 = self._make_layer(96, 6, last_channel)
        self.layer6 = self._make_layer(160, 6, last_channel)

        # building the last several layers
        last_channel = make_divisible(1280 * width_mult, 8) if width_mult > 1.0 else 1280
        self.conv2 = nn.Conv2d(last_channel, last_channel, kernel_size=1, stride=1, padding=0, bias=False)
        self.bn2 = nn.BatchNorm2d(last_channel)
        self.conv3 = nn.Conv2d(last_channel, num_classes, kernel_size=1, stride=1, padding=0, bias=False)
        self.bn3 = nn.BatchNorm2d(num_classes)
        self.avgpool = nn.AdaptiveAvgPool2d((1, 1))

    def _make_layer(self, out_channels, t, last_channel):
        layers = []
        for i in range(t):
            layers.append(InvertedResidual(last_channel, out_channels, 2, 6))
            last_channel = out_channels
        return nn.Sequential(*layers)

    def forward(self, x):
        x = self.conv1(x)
        x = self.bn1(x)
        x = self.relu(x)

        x = self.layer1(x)
        x = self.layer2(x)
        x = self