PyTorch 深度学习 || 4. 自编码网络 | Ch4.1 全连接变分自编码网络 VAE

全连接变分自编码网络 VAE

自编码由3个网络层：输入层、隐藏层和输出层。其中输入层的样本也会充当输出层的角色，即这个神经网络就是一个尽可能复线输入层信号的神经网络，具体结构如下

代码实现

step 1：计算环境和数据的准备

import os
import torch
import torch.nn as nn
import torch.nn.functional as F
import torchvision
from torchvision import transforms
from torchvision.utils import save_image
 
 # 设备配置
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
 
# 如果不存在则创建目录
sample_dir = 'samples'
if not os.path.exists(sample_dir):
    os.makedirs(sample_dir)
 
 # 超参数
image_size = 784
h_dim = 400
z_dim = 20
num_epochs = 15
batch_size = 128
learning_rate = 1e-3
 
# MNIST 数据集
dataset = torchvision.datasets.MNIST(root='./',
                                     train=True,
                                     transform=transforms.ToTensor(),
                                     download=False)
 
# 数据加载器
data_loader = torch.utils.data.DataLoader(dataset=dataset,
                                          batch_size=batch_size, 
                                          shuffle=True)

step 2: 构造全连接的变分自编码网络

# VAE模型
class VAE(nn.Module):
    def __init__(self, image_size=784, h_dim=400, z_dim=20):
        super(VAE, self).__init__()
        self.fc1 = nn.Linear(image_size, h_dim)
        self.fc2 = nn.Linear(h_dim, z_dim)
        self.fc3 = nn.Linear(h_dim, z_dim)
        self.fc4 = nn.Linear(z_dim, h_dim)
        self.fc5 = nn.Linear(h_dim, image_size)
        
    def encode(self, x):
        h = F.relu(self.fc1(x))
        return self.fc2(h), self.fc3(h)
    
    def reparameterize(self, mu, log_var):
        std = torch.exp(log_var/2)
        eps = torch.randn_like(std)
        return mu + eps * std
 
    def decode(self, z):
        h = F.relu(self.fc4(z))
        return F.sigmoid(self.fc5(h))
    
    def forward(self, x):
        mu, log_var = self.encode(x)
        z = self.reparameterize(mu, log_var)
        x_reconst = self.decode(z)
        return x_reconst, mu, log_var


model = VAE().to(device)
optimizer = torch.optim.Adam(model.parameters(), lr=learning_rate)

step 3：训练过程

# 开始训练
for epoch in range(num_epochs):
    for i, (x, _) in enumerate(data_loader):
        # 前传
        x = x.to(device).view(-1, image_size)
        x_reconst, mu, log_var = model(x)
        
        # 计算重建损失和kl散度
        reconst_loss = F.binary_cross_entropy(x_reconst, x, size_average=False)
        kl_div = - 0.5 * torch.sum(1 + log_var - mu.pow(2) - log_var.exp())
        
        # 反向传播和优化
        loss = reconst_loss + kl_div
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()
        
        if (i+1) % 10 == 0:
            print ("Epoch[{}/{}], Step [{}/{}], Reconst Loss: {:.4f}, KL Div: {:.4f}" 
                   .format(epoch+1, num_epochs, i+1, len(data_loader), reconst_loss.item(), kl_div.item()))
    
    with torch.no_grad():
        # 保存采样图像
        z = torch.randn(batch_size, z_dim).to(device)
        out = model.decode(z).view(-1, 1, 28, 28)
        save_image(out, os.path.join(sample_dir, 'sampled-{}.png'.format(epoch+1)))
 
        # 保存重建的图像
        out, _, _ = model(x)
        x_concat = torch.cat([x.view(-1, 1, 28, 28), out.view(-1, 1, 28, 28)], dim=3)
        save_image(x_concat, os.path.join(sample_dir, 'reconst-{}.png'.format(epoch+1)))