版权说明:此文章为本人原创内容,转载请注明出处,谢谢合作!
Pytorch实战1:LeNet手写数字识别 (MNIST数据集)
实验环境:
- Pytorch 0.4.0
- torchvision 0.2.1
- Python 3.6
- CUDA8+cuDNN v7 (可选)
- Win10+Pycharm
Pytorch 0.4.0版本说明
Pyotrch 0.4.0已经支持Windows用户了,其中有多项重大更新,主要包括Tensor/Variable 合并,新增device-agnostic code等等。建议大家安装最新版Pytorch,后面的代码都是基于最新版的,使用老版Pytorch可能会报错。
##LeNet网络结构:
[外链图片转存失败,源站可能有防盗链机制,建议将图片保存下来直接上传(img-s5YyOHZD-1669269967109)(//img-blog.csdn.net/20180426101343001?watermark/2/text/Ly9ibG9nLmNzZG4ubmV0L3N1bnFpYW5kZTg4/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/SouthEast)]
网络参数:
layer name | kernel size | output size |
---|---|---|
输入层 | \ | 32x32 |
卷积层1 | 5x5 | 6x28x28 |
MaxPool1 | 2x2, stride=2 | 6x14x14 |
卷积层2 | 5x5 | 16x10x10 |
MaxPool2 | 2x2, stride=2 | 16x5x5 |
全连接层1 | 16x5x5 | 120 |
全连接层2 | 120 | 84 |
输出层 | 84 | 10 |
Pytorch训练代码:
使用的数据集是MNIST,其中包括6万张28x28的训练样本,1万张测试样本。
import torch
import torchvision as tv
import torchvision.transforms as transforms
import torch.nn as nn
import torch.optim as optim
import argparse
# 定义是否使用GPU
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
# 定义网络结构
class LeNet(nn.Module):
def __init__(self):
super(LeNet, self).__init__()
self.conv1 = nn.Sequential( #input_size=(1*28*28)
nn.Conv2d(1, 6, 5, 1, 2), #padding=2保证输入输出尺寸相同
nn.ReLU(), #input_size=(6*28*28)
nn.MaxPool2d(kernel_size=2, stride=2),#output_size=(6*14*14)
)
self.conv2 = nn.Sequential(
nn.Conv2d(6, 16, 5),
nn.ReLU(), #input_size=(16*10*10)
nn.MaxPool2d(2, 2) #output_size=(16*5*5)
)
self.fc1 = nn.Sequential(
nn.Linear(16 * 5 * 5, 120),
nn.ReLU()
)
self.fc2 = nn.Sequential(
nn.Linear(120, 84),
nn.ReLU()
)
self.fc3 = nn.Linear(84, 10)
# 定义前向传播过程,输入为x
def forward(self, x):
x = self.conv1(x)
x = self.conv2(x)
# nn.Linear()的输入输出都是维度为一的值,所以要把多维度的tensor展平成一维
x = x.view(x.size()[0], -1)
x = self.fc1(x)
x = self.fc2(x)
x = self.fc3(x)
return x
#使得我们能够手动输入命令行参数,就是让风格变得和Linux命令行差不多
parser = argparse.ArgumentParser()
parser.add_argument('--outf', default='./model/', help='folder to output images and model checkpoints') #模型保存路径
parser.add_argument('--net', default='./model/net.pth', help="path to netG (to continue training)") #模型加载路径
opt = parser.parse_args()
# 超参数设置
EPOCH = 8 #遍历数据集次数
BATCH_SIZE = 64 #批处理尺寸(batch_size)
LR = 0.001 #学习率
# 定义数据预处理方式
transform = transforms.ToTensor()
# 定义训练数据集
trainset = tv.datasets.MNIST(
root='./data/',
train=True,
download=True,
transform=transform)
# 定义训练批处理数据
trainloader = torch.utils.data.DataLoader(
trainset,
batch_size=BATCH_SIZE,
shuffle=True,
)
# 定义测试数据集
testset = tv.datasets.MNIST(
root='./data/',
train=False,
download=True,
transform=transform)
# 定义测试批处理数据
testloader = torch.utils.data.DataLoader(
testset,
batch_size=BATCH_SIZE,
shuffle=False,
)
# 定义损失函数loss function 和优化方式(采用SGD)
net = LeNet().to(device)
criterion = nn.CrossEntropyLoss() # 交叉熵损失函数,通常用于多分类问题上
optimizer = optim.SGD(net.parameters(), lr=LR, momentum=0.9)
# 训练
if __name__ == "__main__":
for epoch in range(EPOCH):
sum_loss = 0.0
# 数据读取
for i, data in enumerate(trainloader):
inputs, labels = data
inputs, labels = inputs.to(device), labels.to(device)
# 梯度清零
optimizer.zero_grad()
# forward + backward
outputs = net(inputs)
loss = criterion(outputs, labels)
loss.backward()
optimizer.step()
# 每训练100个batch打印一次平均loss
sum_loss += loss.item()
if i % 100 == 99:
print('[%d, %d] loss: %.03f'
% (epoch + 1, i + 1, sum_loss / 100))
sum_loss = 0.0
# 每跑完一次epoch测试一下准确率
with torch.no_grad():
correct = 0
total = 0
for data in testloader:
images, labels = data
images, labels = images.to(device), labels.to(device)
outputs = net(images)
# 取得分最高的那个类
_, predicted = torch.max(outputs.data, 1)
total += labels.size(0)
correct += (predicted == labels).sum()
print('第%d个epoch的识别准确率为:%d%%' % (epoch + 1, (100 * correct / total)))
torch.save(net.state_dict(), '%s/net_%03d.pth' % (opt.outf, epoch + 1))
#实验结果:
[外链图片转存失败,源站可能有防盗链机制,建议将图片保存下来直接上传(img-xBitzB1Z-1669269967110)(//img-blog.csdn.net/20180426105342789?watermark/2/text/Ly9ibG9nLmNzZG4ubmV0L3N1bnFpYW5kZTg4/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/SouthEast)]