Pytorch实战：用经典网络实现猫狗大战
一：数据集准备： 
数据集来源于研习社：猫狗大战--经典图像分类题 - AI算法竞赛-AI研习社 (yanxishe.com)
二：读取数据：
class Cat_and_Dog_Dataset(Dataset):
def __init__(self, filepath):
self.images = []
self.labels = []
self.transform = transform
for filename in tqdm(os.listdir(filepath)):
image = Image.open(filepath + filename)
image = image.resize((224,224)) #裁剪图片
image = self.transform(image) #转化为Tensor格式
self.images.append(image)
if filename.split('_')[0] == 'cat': #添加标签
self.labels.append(0)
elif filename.split('_')[0] == 'dog':
self.labels.append(1)
self.labels = torch.LongTensor(self.labels) #标签转化位Tensor格式
print(self.labels)
def __getitem__(self, index): #构造迭代器
return self.images[index], self.labels[index]
def __len__(self): #迭代器长度
images = np.array(self.images)
len = images.shape[0]
return len
train_data = Cat_and_Dog_Dataset('cat_dog/train/') #加载训练集
train_loader = DataLoader(dataset=train_data, batch_size=64, shuffle=True)
val_data = Cat_and_Dog_Dataset('cat_dog/val/') #加载验证集
val_loader = DataLoader(dataset=val_data, batch_size=64, shuffle=True)
二：搭建网络：
class InceptionA(torch.nn.Module): #构造Inception层
def __init__(self,in_channels):
super(InceptionA,self).__init__()
self.branch1x1 = torch.nn.Conv2d(in_channels,16,kernel_size = 1)
self.branch5x5_1 = torch.nn.Conv2d(in_channels,16,kernel_size = 1)
self.branch5x5_2 = torch.nn.Conv2d(16,24,kernel_size = 5,padding = 2)
self.branch3x3_1 = torch.nn.Conv2d(in_channels,16,kernel_size = 1)
self.branch3x3_2 = torch.nn.Conv2d(16,24,kernel_size = 3,padding = 1)
self.branch3x3_3 = torch.nn.Conv2d(24,24,kernel_size = 3,padding = 1)
self.branch_pool = torch.nn.Conv2d(in_channels,24,kernel_size = 1)
def forward(self,x):
branch1x1 = self.branch1x1(x)
branch5x5 = self.branch5x5_1(x)
branch5x5 = self.branch5x5_2(branch5x5)
branch3x3 = self.branch3x3_1(x)
branch3x3 = self.branch3x3_2(branch3x3)
branch3x3 = self.branch3x3_3(branch3x3)
branch_pool = F.avg_pool2d(x,kernel_size = 3,stride = 1,padding = 1)
branch_pool = self.branch_pool(branch_pool)
outputs = [branch1x1,branch3x3,branch5x5,branch_pool]
return torch.cat(outputs,dim = 1)
class Net(torch.nn.Module): #构造网络
def __init__(self):
super(Net,self).__init__()
self.conv1 = torch.nn.Conv2d(3,10,kernel_size = 5)
self.conv2 = torch.nn.Conv2d(88,20,kernel_size = 5)
self.incep1 = InceptionA(in_channels = 10)
self.incep2 = InceptionA(in_channels = 20)
self.mp = torch.nn.MaxPool2d(2)
self.fc = torch.nn.Linear(247192,2)
def forward(self,x):
in_size = x.size(0)
x = self.mp(F.relu((self.conv1(x))))
x = self.incep1(x)
x = self.mp(F.relu((self.conv2(x))))
x = self.incep2(x)
x = x.view(in_size,-1)
x = self.fc(x)
return x三：构建损失函数和优化器：
criterion = torch.nn.CrossEntropyLoss() #构造损失函数
optimizer = optim.SGD(model.parameters(),lr = 0.001) #构造优化器四：训练模型：
def train(epoch):
running_loss = 0.0 #训练模型
for batch_idx,data in enumerate(train_loader,0):
inputs,targets = data
inputs, targets = inputs.to(device), targets.to(device)
optimizer.zero_grad() #梯度清零
outputs = model(inputs) #前向传播
loss = criterion(outputs,targets) #计算损失
loss.backward() #逆向传播
optimizer.step() #梯度递进
running_loss += loss.item()
print('train loss: %.3f' % (running_loss/batch_idx))五：验证模型：
def val(): #验证模型精度
correct = 0
total = 0
with torch.no_grad(): #不需要梯度，减少计算量
for data in val_loader:
images, labels = data
images, labels = images.to(device), labels.to(device)
outputs = model(images)
_, predicted = torch.max(outputs.data, dim=1)
total += labels.size(0)
correct += (predicted == labels).sum().item()
print('accuracy on test set: %d %% ' % (100*correct/total))
return correct/total源码：
import torch
import numpy as np
from torch.utils.data import DataLoader,Dataset
from torchvision import transforms
import torchvision
import torch.nn.functional as F
import torch.optim as optim
import os
from tqdm import tqdm
from PIL import Image
import matplotlib.pyplot as plt
batch_size = 64
transform = transforms.Compose([transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))])#设置transform
class Cat_and_Dog_Dataset(Dataset):
def __init__(self, filepath):
self.images = []
self.labels = []
self.transform = transform
for filename in tqdm(os.listdir(filepath)):
image = Image.open(filepath + filename)
image = image.resize((224,224)) #裁剪图片
image = self.transform(image) #转化为Tensor格式
self.images.append(image)
if filename.split('_')[0] == 'cat': #添加标签
self.labels.append(0)
elif filename.split('_')[0] == 'dog':
self.labels.append(1)
self.labels = torch.LongTensor(self.labels) #标签转化位Tensor格式
print(self.labels)
def __getitem__(self, index): #构造迭代器
return self.images[index], self.labels[index]
def __len__(self): #迭代器长度
images = np.array(self.images)
len = images.shape[0]
return len
train_data = Cat_and_Dog_Dataset('cat_dog/train/') #加载训练集
train_loader = DataLoader(dataset=train_data, batch_size=64, shuffle=True)
val_data = Cat_and_Dog_Dataset('cat_dog/val/') #加载验证集
val_loader = DataLoader(dataset=val_data, batch_size=64, shuffle=True)
class InceptionA(torch.nn.Module): #构造Inception层
def __init__(self,in_channels):
super(InceptionA,self).__init__()
self.branch1x1 = torch.nn.Conv2d(in_channels,16,kernel_size = 1)
self.branch5x5_1 = torch.nn.Conv2d(in_channels,16,kernel_size = 1)
self.branch5x5_2 = torch.nn.Conv2d(16,24,kernel_size = 5,padding = 2)
self.branch3x3_1 = torch.nn.Conv2d(in_channels,16,kernel_size = 1)
self.branch3x3_2 = torch.nn.Conv2d(16,24,kernel_size = 3,padding = 1)
self.branch3x3_3 = torch.nn.Conv2d(24,24,kernel_size = 3,padding = 1)
self.branch_pool = torch.nn.Conv2d(in_channels,24,kernel_size = 1)
def forward(self,x):
branch1x1 = self.branch1x1(x)
branch5x5 = self.branch5x5_1(x)
branch5x5 = self.branch5x5_2(branch5x5)
branch3x3 = self.branch3x3_1(x)
branch3x3 = self.branch3x3_2(branch3x3)
branch3x3 = self.branch3x3_3(branch3x3)
branch_pool = F.avg_pool2d(x,kernel_size = 3,stride = 1,padding = 1)
branch_pool = self.branch_pool(branch_pool)
outputs = [branch1x1,branch3x3,branch5x5,branch_pool]
return torch.cat(outputs,dim = 1)
class Net(torch.nn.Module): #构造网络
def __init__(self):
super(Net,self).__init__()
self.conv1 = torch.nn.Conv2d(3,10,kernel_size = 5)
self.conv2 = torch.nn.Conv2d(88,20,kernel_size = 5)
self.incep1 = InceptionA(in_channels = 10)
self.incep2 = InceptionA(in_channels = 20)
self.mp = torch.nn.MaxPool2d(2)
self.fc = torch.nn.Linear(247192,2)
def forward(self,x):
in_size = x.size(0)
x = self.mp(F.relu((self.conv1(x))))
x = self.incep1(x)
x = self.mp(F.relu((self.conv2(x))))
x = self.incep2(x)
x = x.view(in_size,-1)
x = self.fc(x)
return x
model = Net()
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model.to(device) #转换为cuda格式
criterion = torch.nn.CrossEntropyLoss() #构造损失函数
optimizer = optim.SGD(model.parameters(),lr = 0.001) #构造优化器
def train(epoch):
running_loss = 0.0 #训练模型
for batch_idx,data in enumerate(train_loader,0):
inputs,targets = data
inputs, targets = inputs.to(device), targets.to(device)
optimizer.zero_grad() #梯度清零
outputs = model(inputs) #前向传播
loss = criterion(outputs,targets) #计算损失
loss.backward() #逆向传播
optimizer.step() #梯度递进
running_loss += loss.item()
print('train loss: %.3f' % (running_loss/batch_idx))
def val(): #验证模型精度
correct = 0
total = 0
with torch.no_grad(): #不需要梯度，减少计算量
for data in val_loader:
images, labels = data
images, labels = images.to(device), labels.to(device)
outputs = model(images)
_, predicted = torch.max(outputs.data, dim=1)
total += labels.size(0)
correct += (predicted == labels).sum().item()
print('accuracy on test set: %d %% ' % (100*correct/total))
return correct/total
if __name__ == '__main__':
accuracy_list = []
epoch_list = []
for epoch in range(10):
train(epoch)
acc = val()
accuracy_list.append(acc)
epoch_list.append(epoch)
plt.plot(epoch_list,accuracy_list)
plt.xlabel(epoch)
plt.ylabel(accuracy_list)
plt.show()