def train(train_loader, val_loader, class_weights, class_encoding):
print("\nTraining...\n")
num_classes = len(class_encoding)
#write the structure of nn into tensorboard
dataiter = iter(train_loader)
images, labels = dataiter.next()
writer = SummaryWriter('runs/fh_enet_tensorboard')
model = ENet(num_classes)
writer.add_graph(model, images)
writer.close()
# Intialize ENet
#model = ENet(num_classes).to(device)
model.to(device)
# Check if the network architecture is correct
print(model)
# We are going to use the CrossEntropyLoss loss function as it's most
# frequentely used in classification problems with multiple classes which
# fits the problem. This criterion combines LogSoftMax and NLLLoss.
criterion = nn.CrossEntropyLoss(weight=class_weights)
# ENet authors used Adam as the optimizer
optimizer = optim.Adam(model.parameters(),
lr=args.learning_rate,
weight_decay=args.weight_decay)
# Learning rate decay scheduler
lr_updater = lr_scheduler.StepLR(optimizer, args.lr_decay_epochs,
args.lr_decay)
# Evaluation metric
if args.ignore_unlabeled:
ignore_index = list(class_encoding).index('unlabeled')
else:
ignore_index = None
metric = IoU(num_classes, ignore_index=ignore_index)
# Optionally resume from a checkpoint
if args.resume:
model, optimizer, start_epoch, best_miou = utils.load_checkpoint(
model, optimizer, args.save_dir, args.name)
print("Resuming from model: Start epoch = {0} "
"| Best mean IoU = {1:.4f}".format(start_epoch, best_miou))
else:
start_epoch = 0
best_miou = 0
# Start Training
print()
train = Train(model, train_loader, optimizer, criterion, metric, device)
val = Test(model, val_loader, criterion, metric, device)
train_loss = []
train_miou = []
test_loss = []
test_miou = []
for epoch in range(start_epoch, args.epochs):
print(">>>> [Epoch: {0:d}] Training".format(epoch))
epoch_loss, (iou,
miou), loss_iter_result = train.run_epoch(args.print_step)
lr_updater.step()
train_loss.append(epoch_loss)
train_miou.append(miou)
print(">>>> [Epoch: {0:d}] Avg. loss: {1:.4f} | Mean IoU: {2:.4f}".
format(epoch, epoch_loss, miou))
for i in range(len(loss_iter_result)):
writer.add_scalar('train loss iteration', loss_iter_result[i],
epoch * len(loss_iter_result) + i + 1)
if (epoch + 1) % 10 == 0 or epoch + 1 == args.epochs:
print(">>>> [Epoch: {0:d}] Validation".format(epoch))
loss, (iou, miou) = val.run_epoch(args.print_step)
test_loss.append(loss)
test_miou.append(miou)
print(">>>> [Epoch: {0:d}] Avg. loss: {1:.4f} | Mean IoU: {2:.4f}".
format(epoch, loss, miou))
# Print per class IoU on last epoch or if best iou
if epoch + 1 == args.epochs or miou > best_miou:
for key, class_iou in zip(class_encoding.keys(), iou):
print("{0}: {1:.4f}".format(key, class_iou))
# Save the model if it's the best thus far
if miou > best_miou:
print("\nBest model thus far. Saving...\n")
best_miou = miou
utils.save_checkpoint(model, optimizer, epoch + 1, best_miou,
args)
for i in range(len(train_loss)):
writer.add_scalar('train loss epoch', train_loss[i], i + 1)
writer.add_scalar('train miou epoch', train_miou[i], i + 1)
for i in range(len(test_loss)):
writer.add_scalar('test loss epoch', test_loss[i], i + 1)
writer.add_scalar('test miou epoch', test_miou[i], i + 1)
writer.close()
return model
def train(train_loader,
val_loader,
class_weights,
class_encoding,
pretrained="./save/ENet.pt"):
print("\nTraining...\n")
num_classes = len(class_encoding)
model = ENet(num_classes)
if pretrained:
model.load_state_dict(torch.load(pretrained)["state_dict"])
# Intialize ENet
model = model.to(device)
# Check if the network architecture is correct
print(model)
# We are going to use the CrossEntropyLoss loss function as it's most
# frequentely used in classification problems with multiple classes which
# fits the problem. This criterion combines LogSoftMax and NLLLoss.
criterion = nn.CrossEntropyLoss(weight=class_weights)
# ENet authors used Adam as the optimizer
optimizer = optim.Adam(model.parameters(),
lr=args.learning_rate,
weight_decay=args.weight_decay)
# Learning rate decay scheduler
lr_updater = lr_scheduler.StepLR(optimizer, args.lr_decay_epochs,
args.lr_decay)
# Evaluation metric
if args.ignore_unlabeled:
ignore_index = list(class_encoding).index('unlabeled')
else:
ignore_index = None
metric = IoU(num_classes, ignore_index=ignore_index)
# Optionally resume from a checkpoint
if args.resume:
model, optimizer, start_epoch, best_miou = utils.load_checkpoint(
model, optimizer, args.save_dir, args.name)
print("Resuming from model: Start epoch = {0} "
"| Best mean IoU = {1:.4f}".format(start_epoch, best_miou))
else:
start_epoch = 0
best_miou = 0
# Start Training
train = Train(model, train_loader, optimizer, criterion, metric, device)
val = Test(model, val_loader, criterion, metric, device)
for epoch in range(start_epoch, args.epochs):
print(">>>> [Epoch: {0:d}] Training".format(epoch))
lr_updater.step()
epoch_loss, (iou, miou) = train.run_epoch(args.print_step)
print(">>>> [Epoch: {0:d}] Avg. loss: {1:.4f} | Mean IoU: {2:.4f}".
format(epoch, epoch_loss, miou))
if epoch % 10 == 0 or epoch + 1 == args.epochs:
print(">>>> [Epoch: {0:d}] Validation".format(epoch))
loss, (iou, miou) = val.run_epoch(args.print_step)
print(">>>> [Epoch: {0:d}] Avg. loss: {1:.4f} | Mean IoU: {2:.4f}".
format(epoch, loss, miou))
# Print per class IoU on last epoch or if best iou
if epoch + 1 == args.epochs or miou > best_miou:
for key, class_iou in zip(class_encoding.keys(), iou):
print("{0}: {1:.4f}".format(key, class_iou))
# Save the model if it's the best thus far
if miou > best_miou:
print("\nBest model thus far. Saving...\n")
best_miou = miou
utils.save_checkpoint(model, optimizer, epoch + 1, best_miou,
args)
return model