本节使用交叉熵的知识来解决一个多分类问题。
本节所构建的神经网络不再是单层网络
如图是一个十分类问题(十个输出)。
这里先建立三个线性层,
代码语言:javascript复制import torch
import torch.nn.functional as F
# 先建立三个线性层结构
# 建立 784=>200=>200=>10的结构
w1, b1 = torch.randn(200, 784, requires_grad=True),
torch.randn(200, requires_grad=True)
# 之前讲过,括号内分别为(ch_out, ch_in),784是28*28乘积得来,对于常用的mnist数据集,多采用这种像素
w2, b2 = torch.randn(200, 200, requires_grad=True),
torch.randn(200, requires_grad=True)
# 每个层均具有w、b参数
w3, b3 = torch.randn(10, 200, requires_grad=True),
torch.randn(10, requires_grad=True)
# 中间层虽然前后输出维度相同,均是200,但并不是没有作用,而是经历了特征变换的过程
# 进行了[784, 200]=>[200, 200]=>[200, 10]的降维变换
# 将forward过程写进一个函数里面
def forward(x):
x = x@w1.t() b1
# 进行矩阵相乘
x = F.relu(x)
# 使用relu激活函数
x = x@w2.t() b2
x = F.relu(x)
x = x@w3.t() b3
x = F.relu(x)
return x
# 注意 这里返回的x是logits,没有经过sigmoid和softmax
这里完成了tensor的建立和forward过程,下面介绍train(训练)部分。
代码语言:javascript复制# 训练过程首先要建立一个优化器,引入相关工具包
import torch.optim as optim
import torch.nn as nn
learning_rate = 1e-3
optimizer = optim.SGD([w1, b1, w2, b2, w3, b3], lr=learning_rate)
# 这里优化器优化的目标是三种全连接层的变量
criteon = nn.CrossEntropyLoss()
# 这里使用的是crossentropyloss
这里先要求掌握以上代码的书写 后续需会讲解数据读取、结果验证等其他部分代码。
为方便后续讲解,这里先给出全部代码代码
代码语言:javascript复制import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
from torchvision import datasets, transforms
batch_size=200
learning_rate=0.01
epochs=10
train_loader = torch.utils.data.DataLoader(
datasets.MNIST('../data', train=True, download=True,
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))
])),
batch_size=batch_size, shuffle=True)
test_loader = torch.utils.data.DataLoader(
datasets.MNIST('../data', train=False, transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))
])),
batch_size=batch_size, shuffle=True)
w1, b1 = torch.randn(200, 784, requires_grad=True),
torch.zeros(200, requires_grad=True)
w2, b2 = torch.randn(200, 200, requires_grad=True),
torch.zeros(200, requires_grad=True)
w3, b3 = torch.randn(10, 200, requires_grad=True),
torch.zeros(10, requires_grad=True)
torch.nn.init.kaiming_normal_(w1)
torch.nn.init.kaiming_normal_(w2)
torch.nn.init.kaiming_normal_(w3)
def forward(x):
x = x@w1.t() b1
x = F.relu(x)
x = x@w2.t() b2
x = F.relu(x)
x = x@w3.t() b3
x = F.relu(x)
return x
optimizer = optim.SGD([w1, b1, w2, b2, w3, b3], lr=learning_rate)
criteon = nn.CrossEntropyLoss()
for epoch in range(epochs):
for batch_idx, (data, target) in enumerate(train_loader):
data = data.view(-1, 28*28)
logits = forward(data)
loss = criteon(logits, target)
optimizer.zero_grad()
loss.backward()
# print(w1.grad.norm(), w2.grad.norm())
optimizer.step()
if batch_idx % 100 == 0:
print('Train Epoch: {} [{}/{} ({:.0f}%)]tLoss: {:.6f}'.format(
epoch, batch_idx * len(data), len(train_loader.dataset),
100. * batch_idx / len(train_loader), loss.item()))
test_loss = 0
correct = 0
for data, target in test_loader:
data = data.view(-1, 28 * 28)
logits = forward(data)
test_loss = criteon(logits, target).item()
pred = logits.data.max(1)[1]
correct = pred.eq(target.data).sum()
test_loss /= len(test_loader.dataset)
print('nTest set: Average loss: {:.4f}, Accuracy: {}/{} ({:.0f}%)n'.format(
test_loss, correct, len(test_loader.dataset),
100. * correct / len(test_loader.dataset)))