李沐《动手学深度学习》多层感知机python代码实现

一、多层感知机手动实现

# 多层感知机的手动实现
%matplotlib inline
import torch 
from torch import nn
from d2l import torch as d2l

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

num_inputs, num_outputs, num_first_hiddens = 784, 10, 256

W1 = nn.Parameter(
    torch.randn(num_inputs, num_first_hiddens, requires_grad=True)*0.01)

b1 = nn.Parameter(torch.zeros(num_first_hiddens, requires_grad=True))

W2 = nn.Parameter(
    torch.randn(num_first_hiddens, num_outputs, requires_grad=True)*0.01)

b2 = nn.Parameter(torch.zeros(num_outputs, requires_grad=True))

params = [W1, b1, W2, b2]

# relu函数，输入是隐藏层W1*X + b1
def relu(X):
    zero_x = torch.zeros(X.shape)
    return torch.max(X, zero_x)

# 模型函数，输入是数据X
def net(X):
    X = X.reshape((-1, num_inputs))
    H = relu(X@W1 + b1)
    return (H@W2 + b2)

loss = nn.CrossEntropyLoss(reduction='none')

num_epochs = 10
lr = 0.1
updater = torch.optim.SGD(params, lr=lr)

d2l.train_ch3(net, train_iter, test_iter, loss, num_epochs, updater)

二、利用torch框架实现多层感知机

# 多层感知机的简洁实现
%matplotlib inline
import torch
from d2l import torch as d2l
from torch import nn

def init_weights(m):
    if type(m) == nn.Linear:
        nn.init.normal_(m.weight, std=0.01)
        
net = nn.Sequential(nn.Flatten(), # 先把输入数据(1,28,28)展开为(1,784)
                    nn.Linear(784, 256),  # 784输入->256输出的隐藏层
                    nn.ReLU(),  # 对隐藏层的输出再做一个ReLu函数
                    nn.Linear(256, 10)) # 输出层
net.apply(init_weights)

lr = 0.1
batch_size = 256
num_epochs = 10
train_iter, test_iter = d2l.load_data_fashion_mnist(batch_size)

updater = torch.optim.SGD(net.parameters(), lr=lr)
loss = nn.CrossEntropyLoss(reduction='none')

d2l.train_ch3(net, train_iter, test_iter, loss, num_epochs, updater)