请用python写一个基于mindspore的带有两个残差模块的alexnet神经网络

以下是基于MindSpore的带有两个残差模块的AlexNet神经网络的代码：

import mindspore.nn as nn
import mindspore.ops.operations as P
import mindspore.common.dtype as mstype

class AlexNet(nn.Cell):
    def __init__(self, num_classes=10):
        super(AlexNet, self).__init__()
        self.conv1 = nn.Conv2d(in_channels=3, out_channels=64, kernel_size=11, stride=4, pad_mode='valid')
        self.relu1 = nn.ReLU()
        self.pool1 = nn.MaxPool2d(kernel_size=3, stride=2, pad_mode='valid')
        self.conv2 = nn.Conv2d(in_channels=64, out_channels=192, kernel_size=5, pad_mode='same')
        self.relu2 = nn.ReLU()
        self.pool2 = nn.MaxPool2d(kernel_size=3, stride=2, pad_mode='valid')
        self.conv3 = nn.Conv2d(in_channels=192, out_channels=384, kernel_size=3, pad_mode='same')
        self.relu3 = nn.ReLU()
        self.conv4 = nn.Conv2d(in_channels=384, out_channels=256, kernel_size=3, pad_mode='same')
        self.relu4 = nn.ReLU()
        self.conv5 = nn.Conv2d(in_channels=256, out_channels=256, kernel_size=3, pad_mode='same')
        self.relu5 = nn.ReLU()
        self.pool5 = nn.MaxPool2d(kernel_size=3, stride=2, pad_mode='valid')
        self.flatten = nn.Flatten()
        self.fc1 = nn.Dense(in_channels=6*6*256, out_channels=4096)
        self.relu6 = nn.ReLU()
        self.fc2 = nn.Dense(in_channels=4096, out_channels=4096)
        self.relu7 = nn.ReLU()
        self.fc3 = nn.Dense(in_channels=4096, out_channels=num_classes)
        self.softmax = nn.Softmax(axis=1)
        self.add = P.TensorAdd()

    def construct(self, x):
        x = self.pool1(self.relu1(self.conv1(x)))
        x = self.pool2(self.relu2(self.conv2(x)))
        x = self.relu3(self.conv3(x))
        x = self.relu4(self.conv4(x))
        x = self.pool5(self.relu5(self.conv5(x)))
        x = self.flatten(x)
        x = self.relu6(self.fc1(x))
        x = self.relu7(self.fc2(x))
        x = self.fc3(x)
        x1 = x
        x = self.fc3(self.relu7(self.fc2(self.relu6(self.fc1(x)))))
        x = self.add(x, x1)
        return self.softmax(x)

其中，我们在fc1和fc2之间添加了一个残差模块，在fc2之后再添加一个残差模块。这两个残差模块的实现方式都是使用TensorAdd来实现的。