1 初识PyTorch¶

1.1 张量

1.导入pytorch包

In [1]:

import torch

2.创建一个未初始化的5x3张量

In [3]:

x = torch.empty(5, 3)

print(x)

3.创建一个随机初始化的5x3张量

In [4]:

x = torch.rand(5, 3)

print(x)

4.创建一个5x3的0张量，类型为long

In [5]:

x = torch.zeros(5, 3, dtype=torch.long)

print(x)

5.直接从数组创建张量

In [6]:

x = torch.tensor([5.5, 3])

print(x)

6.创建一个5x3的单位张量，类型为double

In [7]:

x = torch.ones(5, 3, dtype=torch.double)

print(x)

7.从已有的张量创建相同维度的新张量，并且重新定义类型为float

In [9]:

x = torch.randn_like(x, dtype=torch.float)

print(x)

8.打印一个张量的维度

In [10]:

print(x.size())

9.将两个张量相加

In [11]:

y = torch.rand(5, 3)

print(x + y)

# 方法二

# print(torch.add(x, y))

# 方法三

#result = torch.empty(5, 3)

#torch.add(x, y, out=result)

#print(result)

# 方法四

#y.add_(x)

#print(y)

10.打印张量的第一列

In [12]:

print(x[:, 1])

11.将一个4x4的张量resize成一个一维张量

In [5]:

x = torch.randn(4, 4)

y = x.view(16)

print(x.size(),y.size())

12.将一个4x4的张量，resize成一个2x8的张量

In [ ]:

y = x.view(2, 8)

print(x.size(),y.size())

# 方法二

z = x.view(-1, 8) # 确定一个维度，-1的维度会被自动计算

print(x.size(),z.size())

13.从张量中取出数字

In [6]:

x = torch.randn(1)

print(x)

print(x.item())

1.2 Numpy的操作

14.将张量装换成numpy数组

In [8]:

a = torch.ones(5)

print(a)

b = a.numpy()

print(b)

15.将张量+1，并观察上题中numpy数组的变化

In [9]:

a.add_(1)

print(a)

print(b)

16.从numpy数组创建张量

In [11]:

import numpy as np

a = np.ones(5)

b = torch.from_numpy(a)

print(a)

print(b)

17.将numpy数组+1并观察上题中张量的变化

In [12]:

np.add(a, 1, out=a)

print(a)

print(b)

2 自动微分

2.1 张量的自动微分

18.新建一个张量，并设置requires_grad=True

In [22]:

x = torch.ones(2, 2, requires_grad=True)

print(x)

tensor([[1., 1.],

[1., 1.]], requires_grad=True)

19.对张量进行任意操作（y = x + 2）

In [4]:

y = x + 2

print(y)

print(y.grad_fn) # y_就多了一个AddBackward的对象_

20.再对y进行任意操作

In [5]:

z = y * y * 3

out = z.mean()

print(z) # z_多了MulBackward的对象_

print(out) # out_多了MeanBackward的对象_

2.2 梯度

21.对out进行反向传播

In [9]:

out.backward()

22.打印梯度d(out)/dx

In [10]:

print(x.grad) #out=0.25*Σ3(x+2)^2

tensor([[4.5000, 4.5000],

[4.5000, 4.5000]])

23.创建一个结果为矢量的计算过程（y=x*2^n）

In [11]:

x = torch.randn(3, requires_grad=True)

y = x * 2

while y.data.norm() < 1000:

y = y * 2

print(y)

tensor([-1178.7739, 1015.1417, 861.7645], grad_fn=)

24.计算v = [0.1, 1.0, 0.0001]处的梯度

In [12]:

v = torch.tensor([0.1, 1.0, 0.0001], dtype=torch.float)

y.backward(v)

print(x.grad)

tensor([1.0240e+02, 1.0240e+03, 1.0240e-01])

25.关闭梯度的功能

In [13]:

print(x.requires_grad)

print((x ** 2).requires_grad)

with torch.no_grad():

print((x ** 2).requires_grad)

# 方法二

# print(x.requires_grad)

# y = x.detach()

# print(y.requires_grad)

# print(x.eq(y).all())

True

True

False

3 神经网络

这部分会实现LeNet5，结构如下所示

3.1 定义网络

In [2]:

import torch

import torch.nn as nn

import torch.nn.functional as F

class Net(nn.Module):

def __init__(self):

super(Net, self).__init__()

# 26.定义①的卷积层，输入为__32x32的图像，卷积核大小5x5卷积核种类__6

self.conv1 = nn.Conv2d(3, 6, 5)

# 27.定义③的卷积层，输入为前一层__6个特征，卷积核大小5x5，卷积核种类__16

self.conv2 = nn.Conv2d(6, 16, 5)

# 28.定义⑤的全链接层，输入为__16*5*5，输出为__120

self.fc1 = nn.Linear(16 * 5 * 5, 120) # 6*6 from image dimension

# 29.定义⑥的全连接层，输入为__120，输出为__84

self.fc2 = nn.Linear(120, 84)

# 30.定义⑥的全连接层，输入为__84，输出为__10

self.fc3 = nn.Linear(84, 10)

def forward(self, x):

# 31.完成__input-S2，先卷积+relu，再2x2下采样

x = F.max_pool2d(F.relu(self.conv1(x)), (2, 2))

# 32.完成__S2-S4，先卷积+relu，再2x2下采样

x = F.max_pool2d(F.relu(self.conv2(x)), 2) #_卷积核方形时，可以只写一个维度_

# 33._将特征向量扁平成行向量_

x = x.view(-1, 16 * 5 * 5)

# 34._使用_fc1+relu

x = F.relu(self.fc1(x))

# 35._使用_fc2+relu

x = F.relu(self.fc2(x))

# 36._使用_fc3

x = self.fc3(x)

return x

net = Net()

print(net)

Net(

(conv1): Conv2d(3, 6, kernel_size=(5, 5), stride=(1, 1))

(conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))

(fc1): Linear(in_features=400, out_features=120, bias=True)

(fc2): Linear(in_features=120, out_features=84, bias=True)

(fc3): Linear(in_features=84, out_features=10, bias=True)

)

37.打印网络的参数

In [15]:

params = list(net.parameters())

# print(params)

print(len(params))

10

38.打印某一层参数的形状

In [16]:

print(params[0].size())

torch.Size([6, 1, 3, 3])

39.随机输入一个向量，查看前向传播输出

In [27]:

input = torch.randn(1, 3, 32, 32)

out = net(input)

print(out)

tensor([[-0.0672, 0.0347, 0.0879, -0.0769, -0.0172, 0.0181, -0.1240, -0.0204,

0.0145, 0.0922]], grad_fn=)

40.将梯度清零

In [25]:

net.zero_grad()

41.随机一个梯度进行反向传播

In [28]:

out.backward(torch.randn(1, 10))

3.2 损失函数

42.用自带的MSELoss()定义损失函数

In [29]:

criterion = nn.MSELoss()

43.随机一个真值，并用随机的输入计算损失

In [32]:

target = torch.randn(10) # 随机真值

target = target.view(1, -1) # 变成行向量

output = net(input) # 用随机输入计算输出

loss = criterion(output, target) # 计算损失

print(loss)

tensor(1.2790, grad_fn=)

44.将梯度初始化，计算上一步中loss的反向传播

In [ ]:

net.zero_grad()

print('conv1.bias.grad before backward')

print(net.conv1.bias.grad)

45.计算43中loss的反向传播

In [34]:

loss.backward()

print('conv1.bias.grad after backward')

print(net.conv1.bias.grad)

conv1.bias.grad before backward

tensor([0., 0., 0., 0., 0., 0.])

conv1.bias.grad after backward

tensor([ 0.0110, 0.0172, 0.0062, -0.0053, 0.0075, 0.0127])

3.3 更新权重

46.定义SGD优化器算法，学习率设置为0.01

In [3]:

import torch.optim as optim

optimizer = optim.SGD(net.parameters(), lr=0.01)

47.使用优化器更新权重

In [36]:

optimizer.zero_grad()

output = net(input)

loss = criterion(output, target)

loss.backward()

# 更新权重

optimizer.step()

4 训练一个分类器

4.1 读取CIFAR10数据，做标准化

48.构造一个transform，将三通道(0,1)区间的数据转换成(-1,1)的数据

In [4]:

import torchvision

import torchvision.transforms as transforms

transform = transforms.Compose(

[transforms.ToTensor(),

transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])

读取数据集，挑几个看看啥样（比较繁琐，随便看看，不做题了）

In [5]:

from torch.utils.data import Dataset, DataLoader

import os

import pickle

batch_size = 4

def load_cifar_batch(filename):

""" load single batch of cifar """

with open(filename, 'rb') as f:

datadict = pickle.load(f, encoding='iso-8859-1')

X = datadict['data']

Y = datadict['labels']

X = X.reshape(10000, 3, 32, 32).transpose(0,2,3,1).astype("uint8")

Y = np.array(Y)

return list(zip(X, Y))

def load_cifar(ROOT):

dataset = []

for b in range(1,6):

f = os.path.join(ROOT, 'data_batch_%d' % (b, ))

batch = load_cifar_batch(f)

dataset.append(batch)

data_train = np.concatenate(dataset)

del batch

data_test = load_cifar_batch(os.path.join(ROOT, 'test_batch'))

return data_train, data_test

class cifar(Dataset):

def __init__(self, root, segmentation='train', transforms=None):

if segmentation == 'train':

self.data = load_cifar(root)[0]

elif segmentation == 'test':

self.data = load_cifar(root)[1]

self.transform = transform

def __getitem__(self, index):

data = self.data[index][0]

if(self.transform):

data = (self.transform(data))

else:

data = (torch.from_numpy(data))

label = self.data[index][1]

return data, label

def __len__(self):

return len(self.data)

In [6]:

trainset = cifar(root = '/home/kesci/input/cifar10', segmentation='train', transforms=transform)

testset = cifar(root = '/home/kesci/input/cifar10', segmentation='test', transforms=transform)

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

testloader = torch.utils.data.DataLoader(testset, batch_size=batch_size,shuffle=False, num_workers=2)

classes = ('plane', 'car', 'bird', 'cat',

'deer', 'dog', 'frog', 'horse', 'ship', 'truck')

In [8]:

import matplotlib.pyplot as plt

import numpy as np

def imshow(img):

img = img / 2 + 0.5 # 把数据退回(0,1)区间

npimg = img.numpy()

plt.imshow(np.transpose(npimg, (1, 2, 0)))

plt.show()

# 随机取一些数据

dataiter = iter(trainloader)

images, labels = dataiter.next()

# 展示图片

imshow(torchvision.utils.make_grid(images))

# 展示分类

print(' '.join('%5s' % classes[labels[j]] for j in range(4)))

frog dog frog deer

4.2 建立网络

这部分沿用前面的网络

In [9]:

net2 = Net()

4.3 定义损失函数和优化器

49.定义交叉熵损失函数

In [10]:

criterion2 = nn.CrossEntropyLoss()

50.定义SGD优化器算法，学习率设置为0.001，momentum=0.9

In [13]:

optimizer2 = optim.SGD(net2.parameters(), lr=0.001, momentum=0.9)

4.4训练网络

In [14]:

for epoch in range(2):

running_loss = 0.0

for i, data in enumerate(trainloader, 0):

# 获取X,y对

inputs, labels = data

# 51._初始化梯度_

optimizer2.zero_grad()

# 52._前馈_

outputs = net2(inputs)

# 53._计算损失_

loss = criterion2(outputs, labels)

# 54._计算梯度_

loss.backward()

# 55._更新权值_

optimizer2.step()

# 每2000个数据打印平均代价函数值

running_loss += loss.item()

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')

[1, 2000] loss: 2.178

[1, 4000] loss: 1.835

[1, 6000] loss: 1.678

[1, 8000] loss: 1.585

[1, 10000] loss: 1.537

[1, 12000] loss: 1.466

[2, 2000] loss: 1.423

[2, 4000] loss: 1.371

[2, 6000] loss: 1.363

[2, 8000] loss: 1.351

[2, 10000] loss: 1.317

[2, 12000] loss: 1.327

Finished Training

4.5 使用模型预测

取一些数据

In [17]:

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)))

GroundTruth: cat ship ship plane

56.使用模型预测

In [18]:

outputs = net2(images)

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

print('Predicted: ', ' '.join('%5s' % classes[predicted[j]]

for j in range(4)))

Predicted: cat ship ship ship

57.在测试集上进行打分

In [19]:

correct = 0

total = 0

with torch.no_grad():

for data in testloader:

images, labels = data

outputs = net2(images)

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

total += labels.size(0)

correct += (predicted == labels).sum().item()

print('Accuracy of the network on the 10000 test images: %d %%' % (

100 * correct / total))

Accuracy of the network on the 10000 test images: 52 %

4.6 存取模型

58.保存训练好的模型

In [21]:

PATH = './cifar_net.pth'

torch.save(net.state_dict(), PATH)

59.读取保存的模型

In [25]:

pretrained_net = torch.load(PATH)

60.加载模型

In [27]:

net3 = Net()

net3.load_state_dict(pretrained_net)

Out[27]: