3.8 多层感知机

3.8.1 从零开始实现

3.8.1.1 获取和读取数据

batch_size = 256
train_iter, test_iter = d2l.load_data_fashion_mnist(batch_size)

3.8.1.2 定义模型参数

Mxnet

num_inputs, num_outputs, num_hiddens = 784, 10, 256

W1 = nd.random.normal(scale=0.01, shape=(num_inputs, num_hiddens))
b1 = nd.zeros(num_hiddens)
W2 = nd.random.normal(scale=0.01, shape=(num_hiddens, num_outputs))
b2 = nd.zeros(num_outputs)
params = [W1, b1, W2, b2]

for param in params:
    param.attach_grad()

Pytorch

num_inputs, num_outputs, num_hiddens = 784, 10, 256

W1 = torch.tensor(np.random.normal(0, 0.01, (num_inputs, num_hiddens)), dtype=torch.float)
b1 = torch.zeros(num_hiddens, dtype=torch.float)
W2 = torch.tensor(np.random.normal(0, 0.01, (num_hiddens, num_outputs)), dtype=torch.float)
b2 = torch.zeros(num_outputs, dtype=torch.float)

params = [W1, b1, W2, b2]
for param in params:
    param.requires_grad_(requires_grad=True)

3.8.1.3 定义激活函数

Mxnet

def relu(X):
    return nd.maximum(X, 0)

Pytorch

def relu(X):
    return torch.max(input=X, other=torch.tensor(0.0))

3.8.1.4 定义模型

Mxnet

def net(X):
    X = X.reshape((-1, num_inputs))
    H = relu(nd.dot(X, W1) + b1)
    return nd.dot(H, W2) + b2

Pytorch

def net(X):
    X = X.view((-1, num_inputs))
    H = relu(torch.matmul(X, W1) + b1)
    return torch.matmul(H, W2) + b2

3.8.1.5 定义损失函数

Mxnet

loss = gloss.SoftmaxCrossEntropyLoss()

Pytorch

loss = torch.nn.CrossEntropyLoss()

3.8.1.6 训练模型

num_epochs, lr = 5, 0.5
d2l.train_ch3(net, train_iter, test_iter, loss, num_epochs, batch_size,
              params, lr)

3.8.2 简洁实现

3.8.2.1 定义模型

Mxnet

net = nn.Sequential()
net.add(nn.Dense(256, activation='relu'),
        nn.Dense(10))
net.initialize(init.Normal(sigma=0.01))

Pytorch

num_inputs, num_outputs, num_hiddens = 784, 10, 256
    
net = nn.Sequential(
        d2l.FlattenLayer(),
        nn.Linear(num_inputs, num_hiddens),
        nn.ReLU(),
        nn.Linear(num_hiddens, num_outputs), 
        )
    
for params in net.parameters():
    init.normal_(params, mean=0, std=0.01)

3.8.2.2 读取数据并训练模型

Mxnet

batch_size = 256
train_iter, test_iter = d2l.load_data_fashion_mnist(batch_size)

loss = gloss.SoftmaxCrossEntropyLoss()
trainer = gluon.Trainer(net.collect_params(), 'sgd', {'learning_rate': 0.5})
num_epochs = 5
d2l.train_ch3(net, train_iter, test_iter, loss, num_epochs, batch_size, None,
              None, trainer)

Pytorch

batch_size = 256
train_iter, test_iter = d2l.load_data_fashion_mnist(batch_size)
loss = torch.nn.CrossEntropyLoss()

optimizer = torch.optim.SGD(net.parameters(), lr=0.5)

num_epochs = 5
d2l.train_ch3(net, train_iter, test_iter, loss, num_epochs, batch_size, None, None, optimizer)

来源：CSDN

作者：咕噜呱啦

链接：https://blog.csdn.net/qinhuiqiao/article/details/104317121