PyTorch深度学习实践10——卷积神经网络基础

• 卷积过程示意图：
• 卷积核的数量要和输入的通道数（Channel数）相等
• N个输入channel，1个输出channel：
•  N个输入channel，M个输出channel：
• 构造一层卷积层（4维张量）需要四个维度：输入大小，输出大小，卷积核W，卷积核H

构造卷积层实例代码：

import torch

in_channels, out_channels = 5, 10
width, height = 100, 100
kernel_size = 3
batch_size = 1

inputs = torch.randn(batch_size,
                     in_channels,
                     width,
                     height)

conv_layer = torch.nn.Conv2d(in_channels,
                             out_channels,
                             kernel_size=kernel_size)

outputs = conv_layer(inputs)
print(inputs.shape)
print(outputs.shape)
print(conv_layer.weight.shape)

•  卷积层的其他参数：填充Padding；padding的目的：保持输出的尺寸不会变小
• 一点理解：输入尺寸是100x100，卷积核尺寸3x3，则输出会是98x98；若卷积核大小是5x5，则输出会是96x96，即width，height都会减小3/2=1和5/2=2。结合卷积时的对应相乘关系即可理解。
• 如上图，卷积核尺寸是3x3，input中的红框中心4和卷积核中心5对应，则output的尺寸会减少两行和两列
• 卷积层的其他参数：步长Stride；目的：有效减少featuremap的尺寸
• 一个下采样方法：MaxPooling（最大池化）；2x2的最大池化层，即stride=2

 一个简单的卷积神经网络（注意观察size的变化）

课上代码：

import torch
import torch.nn as nn
from torchvision import transforms
from torchvision import datasets
from torch.utils.data import DataLoader
import torch.nn.functional as F
import torch.optim as optim
import time

# prepare dataset

batch_size = 64
transform = transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.1307,), (0.3081,))])

train_dataset = datasets.MNIST(root='../dataset/mnist/', train=True, download=True, transform=transform)
train_loader = DataLoader(train_dataset, shuffle=True, batch_size=batch_size)
test_dataset = datasets.MNIST(root='../dataset/mnist/', train=False, download=True, transform=transform)
test_loader = DataLoader(test_dataset, shuffle=False, batch_size=batch_size)


# design model using class

class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 10, kernel_size=5)
        self.conv2 = nn.Conv2d(10, 20, kernel_size=5)
        self.pooling = nn.MaxPool2d(2)
        self.fc = nn.Linear(320, 10)

    def forward(self, x):
        # flatten data from (n,1,28,28) to (n, 784)
        batch_size = x.size(0)
        x = F.relu(self.pooling(self.conv1(x)))
        x = F.relu(self.pooling(self.conv2(x)))
        x = x.view(batch_size, -1)
        x = self.fc(x)
        return x

model = Net()

# construct loss and optimizer
criterion = torch.nn.CrossEntropyLoss()
optimizer = optim.SGD(model.parameters(), lr=0.01, momentum=0.5)


# training cycle forward, backward, update


def train(epoch):
    running_loss = 0.0
    for batch_idx, data in enumerate(train_loader, 0):
        inputs, target = data
        optimizer.zero_grad()

        outputs = model(inputs)
        loss = criterion(outputs, target)
        loss.backward()
        optimizer.step()

        running_loss += loss.item()
        if batch_idx % 300 == 299:
            print('[%d, %5d] loss: %.3f' % (epoch + 1, batch_idx + 1, running_loss / 300))
            running_loss = 0.0


def test():
    correct = 0
    total = 0
    with torch.no_grad():
        for data in test_loader:
            images, labels = data
            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))


if __name__ == '__main__':
    torch.cuda.synchronize()
    start = time.time()
    for epoch in range(10):
        train(epoch)
        test()
    torch.cuda.synchronize()
    end = time.time()
    time_elapsed = end - start
    print('Training complete in {:.0f}m {:.0f}s'.format(
        time_elapsed // 60, time_elapsed % 60))

# Training complete in 3m 50s

如何把运算迁移到GPU上： 

1. Move Model to GPU（定义device，把模型参数和显存转换成CUDA Tensor）
2. Move Tensors to GPU（在train和test函数中把输入扔进GPU即可）

GPU版本代码：

import torch
import torch.nn as nn
from torchvision import transforms
from torchvision import datasets
from torch.utils.data import DataLoader
import torch.nn.functional as F
import torch.optim as optim
import time

# prepare dataset

batch_size = 64
transform = transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.1307,), (0.3081,))])

train_dataset = datasets.MNIST(root='../dataset/mnist/', train=True, download=True, transform=transform)
train_loader = DataLoader(train_dataset, shuffle=True, batch_size=batch_size)
test_dataset = datasets.MNIST(root='../dataset/mnist/', train=False, download=True, transform=transform)
test_loader = DataLoader(test_dataset, shuffle=False, batch_size=batch_size)


# design model using class

class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 10, kernel_size=5)
        self.conv2 = nn.Conv2d(10, 20, kernel_size=5)
        self.pooling = nn.MaxPool2d(2)
        self.fc = nn.Linear(320, 10)

    def forward(self, x):
        # flatten data from (n,1,28,28) to (n, 784)
        batch_size = x.size(0)
        x = F.relu(self.pooling(self.conv1(x)))
        x = F.relu(self.pooling(self.conv2(x)))
        x = x.view(batch_size, -1)
        x = self.fc(x)
        return x

model = Net()
device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
model.to(device)

# construct loss and optimizer
criterion = torch.nn.CrossEntropyLoss()
optimizer = optim.SGD(model.parameters(), lr=0.01, momentum=0.5)


# training cycle forward, backward, update


def train(epoch):
    running_loss = 0.0
    for batch_idx, data in enumerate(train_loader, 0):
        inputs, target = data
        inputs, target = inputs.to(device), target.to(device)

        optimizer.zero_grad()
        outputs = model(inputs)
        loss = criterion(outputs, target)
        loss.backward()
        optimizer.step()

        running_loss += loss.item()
        if batch_idx % 300 == 299:
            print('[%d, %5d] loss: %.3f' % (epoch + 1, batch_idx + 1, running_loss / 300))
            running_loss = 0.0


def test():
    correct = 0
    total = 0
    with torch.no_grad():
        for data in test_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))


if __name__ == '__main__':
    torch.cuda.synchronize()
    start = time.time()
    for epoch in range(10):
        train(epoch)
        test()
    torch.cuda.synchronize()
    end = time.time()
    time_elapsed = end - start
    print('Training complete in {:.0f}m {:.0f}s'.format(
        time_elapsed // 60, time_elapsed % 60))

# Training complete in 2m 44s

课后作业：自己定义新的卷积神经网络：

class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 10, kernel_size=5)
        self.conv2 = nn.Conv2d(10, 20, kernel_size=5)
        self.conv3 = nn.Conv2d(20, 30, kernel_size=3)
        self.pooling = nn.MaxPool2d(2)
        self.fc1 = nn.Linear(30, 16)
        self.fc2 = nn.Linear(16, 10)

    def forward(self, x):
        # flatten data from (n,1,28,28) to (n, 784)
        batch_size = x.size(0)
        x = F.relu(self.pooling(self.conv1(x)))
        x = F.relu(self.pooling(self.conv2(x)))
        x = F.relu(self.pooling(self.conv3(x)))
        x = x.view(batch_size, -1)
        x = self.fc1(x)
        x = self.fc2(x)
        return x