我复现的第一个神经网络: LeNet

目录

1. LeNet简介

2. LeNet实现

3. 实验结果

Reference

学习深度学习已经有小一年的时间，看了很多视频和书本内容，学习了很多代码，可始终感觉认知不够扎实。结合李沐老师的视频课程，我决定在本博客中介绍下复现LeNet的过程。代码基于pycharm2021平台，选用python3.8版本+pytorch1.12.1+cu116。基本上把各个包的版本都刷到最新版本，以方便后续的网络升级和向后兼容。

---

1. LeNet简介

LeNet网络 [1] 由时任AT&T贝尔实验室的研究员Yann LeCun提出，可以被看作是卷积神经网络的开山之作。之所以选用LeNet作为尝试复现的第一个神经网络，是因为该网络本身的结构简单清晰，便于理解。作为早期成功应用于银行和邮政系统的实用型卷积神经网络，LeNet的结构足够经典，其中很多思想传承至今。因此，LeNet作为深度网络代码复现的一个经典案例，十分恰当。

我们首先回顾下LeNet的基本结构。输入是一个32*32的单通道图片 (更新版本的minist数据集的图片尺寸可能减到28*28，那么在卷积的时候需要padding以保证卷积后的特征图为28*28)，之后使用一个卷积层，变换出6通道的28*28的C1 feature map；加一步pooling，由28*28的feature压到14*14。之后按照相同的步骤，压出一个16通道的5*5的feature map，最后加两个全连接层，并输出10个元素组成的向量，以判断输入数字的类别。可以看到，整个结构是非常清晰，便于理解的。

---

2. LeNet实现

LeNet的网络搭建如下：

import torch
from torch import nn
from d2l import torch as s2l

class Reshape(torch.nn.Module):
    def forward(self, x):
        return x.view(-1,1,28,28)

net = torch.nn.Sequential(Reshape(),
                          nn.Conv2d(1,6,kernel_size = 5,padding=2),
                          nn.Sigmoid(),
                          nn.AvgPool2d(kernel_size=2,stride=2),
                          nn.Conv2d(6,16,kernel_size=5),
                          nn.Sigmoid(),
                          nn.AvgPool2d(kernel_size=2,stride=2),
                          nn.Flatten(),
                          nn.Linear(16*5*5, 120),
                          nn.Sigmoid(),
                          nn.Linear(120, 84),
                          nn.Sigmoid(),
                          nn.Linear(84, 10))

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)

可以看到，整个网络的实现还是比较简单的。这里，按照李沐老师视频的介绍，我们给一个随机的输入，来输出网络中各个层对于输入数据的改变，结果如下：

Reshape output shape:        torch.Size([1, 1, 28, 28])
Conv2d output shape:          torch.Size([1, 6, 28, 28])
Sigmoid output shape:         torch.Size([1, 6, 28, 28])
AvgPool2d output shape:     torch.Size([1, 6, 14, 14])
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])
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])

在确定网络结构后，我们提取测试数据。这里，我们使用Fashion-MNIST数据集来训练和测试网络的性能。数据提取代码如下：

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

使用GPU计算模型在数据集上的精度：

def evaluate_accuracy_gpu(net, data_iter,device=None):
    if isinstance(net, torch.nn.Module):
        net.eval()
        if not device:
            device = next(iter(net.parameters())).device
    metric = d2l.Accumulator(2)
    for X,y in data_iter:
        if isinstance(X,list):
            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]

添加训练函数的完整代码：

import torch
from torch import nn
from d2l import torch as d2l

class Reshape(torch.nn.Module):
    def forward(self, x):
        return x.view(-1,1,28,28)

def evaluate_accuracy_gpu(net, data_iter,device=None):
    if isinstance(net, torch.nn.Module):
        net.eval()
        if not device:
            device = next(iter(net.parameters())).device
    metric = d2l.Accumulator(2)
    for X,y in data_iter:
        if isinstance(X,list):
            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):#lr: learning rate
    """train a model woth GPU"""
    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):
        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()
            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))
        test_acc = evaluate_accuracy_gpu(net, test_iter)
        animator.add(epoch + 1, (None, None, test_acc))
        print('Epoch:', epoch)
        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)}')
    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)}')
    print('finished')


def main():
    net = torch.nn.Sequential(Reshape(),
                              nn.Conv2d(1, 6, kernel_size=5, padding=2),
                              nn.Sigmoid(),
                              nn.AvgPool2d(kernel_size=2, stride=2),
                              nn.Conv2d(6, 16, kernel_size=5),
                              nn.Sigmoid(),
                              nn.AvgPool2d(kernel_size=2, stride=2),
                              nn.Flatten(),
                              nn.Linear(16 * 5 * 5, 120),
                              nn.Sigmoid(),
                              nn.Linear(120, 84),
                              nn.Sigmoid(),
                              nn.Linear(84, 10))
    batch_size = 256
    train_iter, test_iter = d2l.load_data_fashion_mnist(batch_size=batch_size)
    lr, num_epochs = 0.9, 10
    train_ch6(net, train_iter, test_iter, num_epochs, lr, d2l.try_gpu())

if __name__ == '__main__':
    main()

---

3. 实验结果

打印结果：

training on cuda:0
Epoch: 0
loss 2.317, train acc 0.102,test acc 0.100
174566.9 examples/secon cuda:0
Epoch: 1
loss 1.383, train acc 0.459,test acc 0.580
139471.3 examples/secon cuda:0
Epoch: 2
loss 0.857, train acc 0.661,test acc 0.652
115809.0 examples/secon cuda:0
Epoch: 3
loss 0.718, train acc 0.716,test acc 0.701
99568.9 examples/secon cuda:0
Epoch: 4
loss 0.648, train acc 0.748,test acc 0.752
87336.1 examples/secon cuda:0
Epoch: 5
loss 0.590, train acc 0.770,test acc 0.776
77399.1 examples/secon cuda:0
Epoch: 6
loss 0.550, train acc 0.787,test acc 0.781
69605.1 examples/secon cuda:0
Epoch: 7
loss 0.515, train acc 0.800,test acc 0.793
63230.5 examples/secon cuda:0
Epoch: 8
loss 0.485, train acc 0.816,test acc 0.799
57836.1 examples/secon cuda:0
Epoch: 9
loss 0.459, train acc 0.829,test acc 0.761
53456.0 examples/secon cuda:0
loss 0.459, train acc 0.829,test acc 0.761
53456.0 examples/secon cuda:0

动态曲线图： 

注：如果动画无法显示，参考博客：无法显示动图怎么办？

---

Reference

[1] LeCun Y, Bottou L, Bengio Y, et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11): 2278-2324.