def train(train_loader,
model,
m,
criterion,
optimizer,
attack,
device,
epoch,
callback=None):
model.train()
cudnn.benchmark = True
length = len(train_loader)
batch_time = utils.AverageMeter()
losses = utils.AverageMeter()
losses_adv = utils.AverageMeter()
top1 = utils.AverageMeter()
top5 = utils.AverageMeter()
adv_acc = utils.AverageMeter()
l2_adv = utils.AverageMeter()
end = time.time()
for i, (data, labels) in enumerate(train_loader):
data = data.to(device)
labels = labels.to(device, non_blocking=True)
if args.adv and epoch >= args.start_adv_epoch:
model.eval()
utils.requires_grad_(m, False)
clean_logits = model(data)
loss = criterion(clean_logits, labels)
adv = attack.attack(model, data, labels)
l2_norms = (adv - data).view(args.batch_size, -1).norm(2, 1)
mean_norm = l2_norms.mean()
if args.max_norm:
adv = torch.renorm(
adv - data, p=2, dim=0, maxnorm=args.max_norm) + data
l2_adv.append(mean_norm.item())
utils.requires_grad_(m, True)
model.train()
adv_logits = model(adv.detach())
loss_adv = criterion(adv_logits, labels)
loss_to_optimize = loss_adv
losses_adv.append(loss_adv.item())
l2_adv.append((adv - data).view(args.batch_size,
-1).norm(p=2, dim=1).mean().item())
adv_acc.append((adv_logits.argmax(1) == labels).sum().item() /
args.batch_size)
else:
clean_logits = model(data)
loss = criterion(clean_logits, labels)
loss_to_optimize = loss
optimizer.zero_grad()
loss_to_optimize.backward()
optimizer.step()
# measure accuracy and record loss
prec1, prec5 = utils.accuracy(clean_logits, labels, topk=(1, 5))
losses.append(loss.item())
top1.append(prec1)
top5.append(prec5)
# measure elapsed time
batch_time.append(time.time() - end)
end = time.time()
if (i + 1) % args.print_freq == 0 or (i + 1) == length:
if args.adv and epoch >= args.start_adv_epoch:
print(
'Epoch: [{0:>2d}][{1:>3d}/{2:>3d}] Time {batch_time.last_avg:.3f}'
'\tLoss {loss.last_avg:.4f}\tAdv {loss_adv.last_avg:.4f}'
'\tPrec@1 {top1.last_avg:.3%}\tPrec@5 {top5.last_avg:.3%}'.
format(epoch,
i + 1,
len(train_loader),
batch_time=batch_time,
loss=losses,
loss_adv=losses_adv,
top1=top1,
top5=top5))
else:
print(
'Epoch: [{0:>2d}][{1:>3d}/{2:>3d}] Time {batch_time.last_avg:.3f}\tLoss {loss.last_avg:.4f}'
'\tPrec@1 {top1.last_avg:.3%}\tPrec@5 {top5.last_avg:.3%}'.
format(epoch,
i + 1,
len(train_loader),
batch_time=batch_time,
loss=losses,
top1=top1,
top5=top5))
if callback:
if args.adv and epoch >= args.start_adv_epoch:
callback.scalars(['train_loss', 'adv_loss'],
i / length + epoch,
[losses.last_avg, losses_adv.last_avg])
callback.scalars(
['train_prec@1', 'train_prec@5', 'adv_acc'],
i / length + epoch, [
top1.last_avg * 100, top5.last_avg * 100,
adv_acc.last_avg * 100
])
callback.scalar('adv_l2', i / length + epoch,
l2_adv.last_avg)
else:
callback.scalar('train_loss', i / length + epoch,
losses.last_avg)
callback.scalars(
['train_prec@1', 'train_prec@5'], i / length + epoch,
[top1.last_avg * 100, top5.last_avg * 100])
lr=args.lr,
momentum=args.momentum,
weight_decay=args.weight_decay)
scheduler = lr_scheduler.StepLR(optimizer,
step_size=args.lr_step,
gamma=args.lr_decay)
attacker = DDN(steps=args.steps, device=DEVICE)
best_acc = 0
best_epoch = 0
for epoch in range(args.epochs):
cudnn.benchmark = True
model.train()
requires_grad_(model, True)
accs = AverageMeter()
losses = AverageMeter()
attack_norms = AverageMeter()
scheduler.step()
length = len(train_loader)
for i, (images, labels) in enumerate(tqdm.tqdm(train_loader, ncols=80)):
images, labels = images.to(DEVICE), labels.to(DEVICE)
if args.adv is not None and epoch >= args.adv:
model.eval()
requires_grad_(model, False)
with torch.no_grad():
accs.append(
(model(images).argmax(1) == labels).float().mean().item())
max_loss = torch.log(torch.tensor(10.)).item() # for callback
best_acc = 0
best_epoch = 0
valacc_final = 0
max_loss = torch.log(torch.tensor(1000.)).item() # for callback
best_acc = 0
best_epoch = 0
for epoch in range(args.epochs):
scheduler.step()
cudnn.benchmark = True
model.train()
requires_grad_(m, True)
accs = AverageMeter()
losses = AverageMeter()
attack_norms = AverageMeter()
# print("len(train_loader) ", len(train_loader))
length = len(train_loader)
widgets = ['train :', Percentage(), ' ', Bar('#'), ' ', Timer(),
' ', ETA(), ' ', FileTransferSpeed()]
pbar = ProgressBar(widgets=widgets)
for i,batch_data in enumerate(tqdm.tqdm(train_loader, ncols=80)):
images, labels =batch_data['image'],batch_data['label_idx'].to(DEVICE)
images = _jpeg_compression3(images).to(DEVICE)
# imh_cam = images[0]
# imh_cam = imh_cam.cpu().numpy() # FloatTensor转为ndarray
# imh_cam = np.transpose(imh_cam, (1, 2, 0)) # 把channel那一维放到最后
# plt.imshow(imh_cam)