shuffle=True,
collate_fn=data.image_label_list_of_masks_collate_function)
validation_dataset_fid = DataLoader(
data.Places365(path_to_index_file=args.path_to_places365,
index_file_name='val.txt',
max_length=6000,
validation=True),
batch_size=args.batch_size,
num_workers=args.batch_size,
shuffle=False,
collate_fn=data.image_label_list_of_masks_collate_function)
validation_dataset = data.Places365(
path_to_index_file=args.path_to_places365, index_file_name='val.txt')
# Init model wrapper
model_wrapper = ModelWrapper(
generator=generator,
discriminator=discriminator,
vgg16=vgg16,
training_dataset=training_dataset,
validation_dataset=validation_dataset,
validation_dataset_fid=validation_dataset_fid,
generator_optimizer=generator_optimizer,
discriminator_optimizer=discriminator_optimizer)
# Perform training
if args.train:
model_wrapper.train(epochs=args.epochs, device=args.device)
# Perform testing
if args.test:
print('FID=', model_wrapper.validate(device=args.device))
model_wrapper.inference(device=args.device)
# Model wrapper
model_wrapper = ModelWrapper(detr=detr,
detr_optimizer=detr_optimizer,
detr_segmentation_optimizer=detr_segmentation_optimizer,
training_dataset=training_dataset,
validation_dataset=validation_dataset,
test_dataset=test_dataset,
loss_function=InstanceSegmentationLoss(
segmentation_loss=SegmentationLoss(),
ohem=args.ohem,
ohem_faction=args.ohem_fraction),
device=device)
# for im, mask,bb,label in training_dataset:
# print(im)
# Perform training
if args.train:
model_wrapper.train(epochs=args.epochs,
optimize_only_segmentation_head_after_epoch=args.only_train_segmentation_head_after_epoch)
# Perform validation
if args.val:
model_wrapper.validate(number_of_plots=30)
# Perform testing
if args.test:
model_wrapper.test()
# Perform testing
if args.inference:
model_wrapper.inference()