P23 LeNet经典神经网络模型
LeNet(LeNet-5) 由两个部分组成：卷积编码器和全连接层密集块
import torch
from matplotlib import pyplot as plt
from torch import nn
from d2l import torch as d2l
class Reshape(torch.nn.Module): #原始输入是32*32
def forward(self,x):
return x.view(-1,1,28,28)
#28*28的图片net = torch.nn.Sequential(
Reshape(),nn.Conv2d(1, 6, kernel_size = 5, padding=2 ),nn.Sigmoid(),
#kernel的窗口是5*5，因为原始数据是32*32，当它变成28*28的时候，等于少了边缘，这里补上，padding=2
#在卷积后加入sigmoid函数
nn.AvgPool2d(kernel_size=2, stride=2), #用平均池化
nn.Conv2d(6, 16, kernel_size=5), nn.Sigmoid(),
#卷积层输入是6，kernel还是5，这里不作padding，结束后继续做sigmoid
nn.AvgPool2d(kernel_size=2, stride=2), nn.Flatten(),
#接上平均池化，和上一层一样，但是最后要把输出变成1维的向量，所以用flatten
nn.Linear(16*5*5,120),nn.Sigmoid(),
#高和宽在池化后，输出是120,16*5*5是最后一层的输出
nn.Linear(120,84),nn.Sigmoid(),
#把120降到84
nn.Linear (84,10))
#最后降到10，因为他有10个类别
#2个卷积层+1个激活层+池化层验证函数
X = torch.rand(size=(1,1,28,28),dtype=torch.float32)
for layer in net:
X = layer(X)
print(layer.__class__.__name__,'output shape:\t',X.shape)打印结果：和书上的图例一样，代码显示的是书上的图例
Reshape output shape: torch.Size([1, 1, 28, 28]) 第一个是batch size=1， 通道=1，输入大小28*28 Conv2d output shape: torch.Size([1, 6, 28, 28]) 到了这里，通道变成了6 Sigmoid output shape: torch.Size([1, 6, 28, 28]) AvgPool2d output shape: torch.Size([1, 6, 14, 14]) 第一个模块结束，输出是14*14，但是通道变成了6 Conv2d output shape: torch.Size([1, 16, 10, 10]) Sigmoid output shape: torch.Size([1, 16, 10, 10]) AvgPool2d output shape: torch.Size([1, 16, 5, 5]) 第二个模块结束，高和宽被减了3倍，通道数从6变成16，是MLP的思想，慢慢把一个很长的向量往下压，最后输出是5*5 卷积把图片大小渐渐变小，但是通道慢慢变多，把空间信息压缩，把抽出来压缩的信息放在不同通道里 最后通过一个多层感知机模型，能训练到最后的输出，极端一点图片大小会变成1，通道会有几千，最后做全连接输出 Flatten output shape: torch.Size([1, 400]) Linear output shape: torch.Size([1, 120]) Sigmoid output shape: torch.Size([1, 120]) Linear output shape: torch.Size([1, 84]) Sigmoid output shape: torch.Size([1, 84]) Linear output shape: torch.Size([1, 10])
LeNet在Fashion-MNIST上实现
batch_size = 256
train_iter, test_iter =d2l.load_data_fashion_mnist(batch_size=batch_size)def evaluate_accuracy_gpu(net, data_iter, device=None): #@save
"""使用GPU计算模型在数据集上的精度"""
if isinstance(net, nn.Module):
net.eval() # 设置为评估模式
if not device:
device = next(iter(net.parameters())).device
# 正确预测的数量，总预测的数量
metric = d2l.Accumulator(2)
with torch.no_grad():
for X, y in data_iter:
if isinstance(X, list):
# BERT微调所需的（之后将介绍）
X = [x.to(device) for x in X]
else:
X = X.to(device)
y = y.to(device)
metric.add(d2l.accuracy(net(X), y), y.numel())
return metric[0] / metric[1]
def train_ch6(net, train_iter, test_iter, num_epochs, lr, device):
"""用GPU训练模型(在第六章定义)"""
def init_weights(m):
if type(m) == nn.Linear or type(m) == nn.Conv2d:
nn.init.xavier_uniform_(m.weight)
net.apply(init_weights)
print('training on', device)
net.to(device)
optimizer = torch.optim.SGD(net.parameters(), lr=lr)
loss = nn.CrossEntropyLoss()
animator = d2l.Animator(xlabel='epoch', xlim=[1, num_epochs],
legend=['train loss', 'train acc', 'test acc'])
timer, num_batches = d2l.Timer(), len(train_iter)
for epoch in range(num_epochs):
# 训练损失之和，训练准确率之和，样本数
metric = d2l.Accumulator(3)
net.train()
for i, (X, y) in enumerate(train_iter):
timer.start()
optimizer.zero_grad()
X, y = X.to(device), y.to(device)
y_hat = net(X)
l = loss(y_hat, y)
l.backward()
optimizer.step()
with torch.no_grad():
metric.add(l * X.shape[0], d2l.accuracy(y_hat, y), X.shape[0])
timer.stop()
train_l = metric[0] / metric[2]
train_acc = metric[1] / metric[2]
if (i + 1) % (num_batches // 5) == 0 or i == num_batches - 1:
animator.add(epoch + (i + 1) / num_batches,(train_l, train_acc, None))
plt.show()
test_acc = evaluate_accuracy_gpu(net, test_iter)
animator.add(epoch + 1, (None, None, test_acc))
plt.show()
print(f'loss {train_l:.3f}, train acc {train_acc:.3f}, '
f'test acc {test_acc:.3f}')
print(f'{metric[2] * num_epochs / timer.sum():.1f} examples/sec '
f'on {str(device)}')
plt.show()训练和评估LeNet-5模型
lr, num_epochs = 0.9,10
train_ch6(net,train_iter,test_iter,num_epochs,lr,d2l.try_gpu())
plt.show()打印结果：
loss 0.467, train acc 0.825, test acc 0.767 10168.4 examples/sec on cpu
动画不能像视频里那样动态显示