我使用 DCGAN 来合成医学图像（512*512）。然而，目前，DCGAN 太不稳定了。因此，我正在尝试将我的 DCGAN 网络更改为 WGAN。
这个链接是我的 DCGAN 网络的原始代码。
数据和参数
# Root directory for dataset
dataroot = f"./processed/{grade}/{grade}/"
# Number of workers for dataloader
workers = 4
# Batch size during training
batch_size = 128
# Spatial size of training images. All images will be resized to this
# size using a transformer.
image_size = 512
# Number of channels in the training images. For color images this is 3
nc = 3
# Size of z latent vector (i.e. size of generator input)
nz = 100
# Size of feature maps in generator
ngf = 16
# Size of feature maps in discriminator
ndf = 16
# Number of training epochs
num_epochs = 500
# Learning rate for optimizers
lr = 0.0002
# Beta1 hyperparam for Adam optimizers
beta1 = 0.5
# Number of GPUs available. Use 0 for CPU mode.
ngpu = 2
# WGAN clip gradient
clamp_num=0.01
我改变了 weight_init()
def weight_init(m):
# weight_initialization: important for wgan
class_name=m.__class__.__name__
if class_name.find('Conv')!=-1:
m.weight.data.normal_(0,0.02)
elif class_name.find('Norm')!=-1:
m.weight.data.normal_(1.0,0.02)
变化生成器
class Generator(nn.Module):
def __init__(self, ngpu):
super(Generator, self).__init__()
self.ngpu = ngpu
self.main = nn.Sequential(
# input is Z, going into a convolution
nn.ConvTranspose2d(nz, ngf * 64, 4, 1, 0, bias=False),
nn.BatchNorm2d(ngf * 64),
nn.ReLU(True),
# state size. (ngf*64) x 4 x 4
nn.ConvTranspose2d(ngf * 64, ngf * 32, 4, 2, 1, bias=False),
nn.BatchNorm2d(ngf * 32),
nn.ReLU(True),
# state size. (ngf*32) x 8 x 8
nn.ConvTranspose2d(ngf * 32, ngf * 16, 4, 2, 1, bias=False),
nn.BatchNorm2d(ngf * 16),
nn.ReLU(True),
# state size. (ngf*16) x 16 x 16
nn.ConvTranspose2d(ngf * 16, ngf * 8, 4, 2, 1, bias=False),
nn.BatchNorm2d(ngf * 8),
nn.ReLU(True),
# state size. (ngf*8) x 32 x 32
nn.ConvTranspose2d(ngf * 8, ngf * 4, 4, 2, 1, bias=False),
nn.BatchNorm2d(ngf * 4),
nn.ReLU(True),
# state size. (ngf*4) x 64 x 64
nn.ConvTranspose2d(ngf * 4, ngf * 2, 4, 2, 1, bias=False),
nn.BatchNorm2d(ngf * 2),
nn.ReLU(True),
# state size. (ngf*2) x 128 x 128
nn.ConvTranspose2d(ngf * 2, ngf, 4, 2, 1, bias=False),
nn.BatchNorm2d(ngf),
nn.ReLU(True),
# state size. (ngf) x 256 x 256
nn.ConvTranspose2d( ngf, nc, 4, 2, 1, bias=False),
nn.Tanh()
# state size. (nc) x 512 x 512
)
def forward(self, x):
return self.main(x)
和鉴别器
class Discriminator(nn.Module):
def __init__(self, ngpu):
super(Discriminator, self).__init__()
self.ngpu = ngpu
self.main = nn.Sequential(
# input is (nc) x 512 x 512
nn.Conv2d(nc, ndf, 4, 2, 1, bias=False),
nn.LeakyReLU(0.2, inplace=True),
# state size. (ndf) x 256 x 256
nn.Conv2d(ndf, ndf * 2, 4, 2, 1, bias=False),
nn.BatchNorm2d(ndf * 2),
nn.LeakyReLU(0.2, inplace=True),
# state size. (ndf*2) x 128 x 128
nn.Conv2d(ndf * 2, ndf * 4, 4, 2, 1, bias=False),
nn.BatchNorm2d(ndf * 4),
nn.LeakyReLU(0.2, inplace=True),
# state size. (ndf*4) x 64 x 64
nn.Conv2d(ndf * 4, ndf * 8, 4, 2, 1, bias=False),
nn.BatchNorm2d(ndf * 8),
nn.LeakyReLU(0.2, inplace=True),
# state size. (ndf*8) x 32 x 32
nn.Conv2d(ndf * 8, ndf * 16, 4, 2, 1, bias=False),
nn.BatchNorm2d(ndf * 16),
nn.LeakyReLU(0.2, inplace=True),
# state size. (ndf*16) x 16 x 16
nn.Conv2d(ndf * 16, ndf * 32, 4, 2, 1, bias=False),
nn.BatchNorm2d(ndf * 32),
nn.LeakyReLU(0.2, inplace=True),
# state size. (ndf*32) x 8 x 8
nn.Conv2d(ndf * 32, ndf * 64, 4, 2, 1, bias=False),
nn.BatchNorm2d(ndf * 64),
nn.LeakyReLU(0.2, inplace=True),
# state size. (ndf*64) x 4 x 4
nn.Conv2d(ndf * 64, 1, 4, 1, 0, bias=False),
# Modification 1: remove sigmoid
# nn.Sigmoid()
)
def forward(self, x):
return self.main(x)
此外，更改优化器
from torch.optim import RMSprop
# modification 2: Use RMSprop instead of Adam
optimizerD = RMSprop(netD.parameters(),lr=lr )
optimizerG = RMSprop(netG.parameters(),lr=lr )
# modification3: No Log in loss
# criterion = nn.BCELoss()
# Create batch of latent vectors that we will use to visualize
# the progression of the generator
fixed_noise = torch.randn(64, nz, 1, 1, device=device)
# Establish convention for real and fake labels during training
real_label = 1
fake_label = 0
最后，训练代码如下。我猜培训代码有问题。（另外，我没有更改打印部分的任何内容）我希望有人可以帮助我如何更改训练代码以在这个意义上运行 WGAN。
# Training Loop
one=torch.FloatTensor([1]).cuda()
mone=-1*one.cuda()
# Lists to keep track of progress
img_list = []
G_losses = []
D_losses = []
iters = 0
num_epochs = 1000
print("Starting Training Loop...")
# For each epoch
for epoch in range(num_epochs):
# For each batch in the dataloader
for i, data in enumerate(dataloader, 0):
for parm in netD.parameters():
parm.data.clamp_(clamp_num,clamp_num)
############################
# (1) Update D network: maximize log(D(x)) + log(1 - D(G(z)))
###########################
## Train with all-real batch
#print(epoch)
netD.zero_grad()
# Format batch
real_cpu = data[0].to(device)
b_size = real_cpu.size(0)
label = torch.full((b_size,), real_label, device=device).float()
#print(real_cpu.shape)
output = netD(real_cpu).view(-1).float()
# Calculate loss on all-real batc
output.backward(one)
# Calculate gradients for D in backward pass
D_x = output.mean().item()
## Train with all-fake batch
# Generate batch of latent vectors
noise = torch.randn(b_size, nz, 1, 1, device=device)
# Generate fake image batch with G
fake = netG(noise)
label.fill_(fake_label)
#print(fake.detach())
# Classify all fake batch with D
output2 = netD(fake.detach()).view(-1).float()
# Calculate D's loss on the all-fake batch
output2.backward(mone)
# Calculate the gradients for this batch
D_G_z1 = output.mean().item()
# Add the gradients from the all-real and all-fake batches
# Update D
optimizerD.step()
############################
# (2) Update G network: maximize log(D(G(z)))
###########################
netG.zero_grad()
label.fill_(real_label) # fake labels are real for generator cost
# Since we just updated D, perform another forward pass of all-fake batch through D
output2 = netD(fake.detach()).view(-1).float()
#output = netD(fake).view(-1)
# Calculate G's loss based on this output
#errG = criterion(output, label)
# Calculate gradients for G
output2.backward()
D_G_z2 = output2.mean().item()
# Update G
optimizerG.step()
# Output training stats
if i % 1000 == 0:
print('[%d/%d][%d/%d]\tLoss_D: %.4f\tLoss_G: %.4f\tD(x): %.4f\tD(G(z)): %.4f / %.4f'
% (epoch, num_epochs, i, len(dataloader),
errD.item(), errG.item(), D_x, D_G_z1, D_G_z2))
# Save Losses for plotting later
G_losses.append(errG.item())
D_losses.append(errD.item())
# Check how the generator is doing by saving G's output on fixed_noise
if (iters % 500 == 0) or ((epoch == num_epochs-1) and (i == len(dataloader)-1)):
with torch.no_grad():
fake = netG(fixed_noise).detach().cpu()
img_list.append(utils.make_grid(fake, padding=0, normalize=True))
iters += 1