pyTorch——训练第一个分类器要点解读

网络构建

数据加载

* 引入函数库
import torch
import torchvision
import torchvision.transforms as transforms

*将读入的数据进行转化：
transform = transforms.Compose(
[transforms.ToTensor(), ***range [0, 255] -> [0.0,1.0]
transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))]) *数据分布归一化到[-1,1]

*利用torch自带的CIFAR10数据集加载训练集
trainset = torchvision.datasets.CIFAR10(root=’./data’, train=True,
download=True, transform=transform)

*生成batch，其中：
*参数：
dataset：Dataset类型，从其中加载数据
batch_size：int，可选。每个batch加载多少样本
shuffle：bool，可选。为True时表示每个epoch都对数据进行洗牌
sampler：Sampler，可选。从数据集中采样样本的方法。
num_workers：int，可选。加载数据时使用多少子进程。默认值为0，表示在主进程中加载数据。
collate_fn：callable，可选。
pin_memory：bool，可选
drop_last：bool，可选。True表示如果最后剩下不完全的batch,丢弃。False表示不丢弃。

trainloader = torch.utils.data.DataLoader(trainset, batch_size=4,
shuffle=True, num_workers=2)

*加载测试集
testset = torchvision.datasets.CIFAR10(root=’./data’, train=False,
download=True, transform=transform)

*测试集batch
testloader = torch.utils.data.DataLoader(testset, batch_size=4,
shuffle=False, num_workers=2)

*定义类别
classes = (‘plane’, ‘car’, ‘bird’, ‘cat’,
‘deer’, ‘dog’, ‘frog’, ‘horse’, ‘ship’, ‘truck’)

*显示一些训练集中的图片与标签
import matplotlib.pyplot as plt
import numpy as np

def imshow(img):
img = img / 2 + 0.5 # unnormalize
npimg = img.numpy()
plt.imshow(np.transpose(npimg, (1, 2, 0)))

*# get some random training images
dataiter = iter(trainloader)
images, labels = dataiter.next()

*# show images
imshow(torchvision.utils.make_grid(images))

*# print labels
print(’ ‘.join(‘%5s’ % classes[labels[j]] for j in range(4)))

---

定义网络

from torch.autograd import Variable ***Variable是最核心的变量
import torch.nn as nn *神经网络库
import torch.nn.functional as F

*定义网络单元
class Net(nn.Module):

 def __init__(self):
    super(Net, self).__init__()
    self.conv1 = nn.Conv2d(3, 6, 5) //3 input image
                                    // channel, 6 output channels
                                    //5x5 square convolution
    self.pool = nn.MaxPool2d(2, 2)
    self.conv2 = nn.Conv2d(6, 16, 5)
    self.fc1 = nn.Linear(16 * 5 * 5, 120)
    self.fc2 = nn.Linear(120, 84)

def forward(self, x):
    //x --> conv1 --> relu --> pool -->x
    x = self.pool(F.relu(self.conv1(x)))
   //x --> conv2 --> relu -->pool --> x
    x = self.pool(F.relu(self.conv2(x))) 
   //view函数将张量x变形成一维向量形式，总特征数不变，为全连接层做准备
    x = x.view(-1, 16 * 5 * 5) 
    x = F.relu(self.fc1(x))
    x = F.relu(self.fc2(x))
    x = self.fc3(x)
    return x

 net = Net()  

损失函数

***use a Classification Cross-Entropy loss and SGD with momentum

import torch.optim as optim

criterion = nn.CrossEntropyLoss()
optimizer = optim.SGD(net.parameters(), lr=0.001, momentum=0.9)

训练过程

for epoch in range(2):  #全部训练集训练两次：epoch=[0,1]

    running_loss = 0.0 #清空loss
    for i, data in enumerate(trainloader, 0):
        # get the inputs
        inputs, labels = data #trainloader返回：id,image,labels

        # 将inputs于labels装进Variable中   
        #（autograd.Varible[data,grad,creator]）
        inputs, labels = Variable(inputs), Variable(labels)

        # zero the parameter gradients
        optimizer.zero_grad()

        # forward + backward + optimize
        outputs = net(inputs)
        loss = criterion(outputs, labels)
        #back ward to every variable recorded in Variable's grad
        loss.backward()
        optimizer.step() #do SGD

        # print statistics
        running_loss += loss.data[0]
        if i % 2000 == 1999:    # print every 2000 mini-batches
            print('[%d, %5d] loss: %.3f' %
                  (epoch + 1, i + 1, running_loss / 2000))
            running_loss = 0.0

print('Finished Training')

测试过程

dataiter = iter(testloader)
images, labels = dataiter.next()

# print images
imshow(torchvision.utils.make_grid(images))
print('GroundTruth: ', ' '.join('%5s' % classes[labels[j]] for j in range(4)))

outputs = net(Variable(images))

_, predicted = torch.max(outputs.data, 1)

print('Predicted: ', ' '.join('%5s' % classes[predicted[j]]
                              for j in range(4)))

***On the whole dataset

correct = 0
total = 0
for data in testloader:
    images, labels = data
    outputs = net(Variable(images))
    _, predicted = torch.max(outputs.data, 1)
    total += labels.size(0)
    correct += (predicted == labels).sum()

print('Accuracy of the network on the 10000 test images: %d %%' % (
    100 * correct / total))

在GPU上训练

*将网络转到GPU上
net.cuda()
*数据也要在GPU上
inputs, labels = Variable(inputs.cuda()), Variable(labels.cuda())