opts.num_style = 1 if opts.style != '' else opts.num_style
# Setup model and data loader
config['vgg_model_path'] = opts.output_path
if opts.trainer == 'MUNIT':
style_dim = config['gen']['style_dim']
trainer = MUNIT_Trainer(config)
else:
sys.exit("Only support MUNIT")
try:
state_dict = torch.load(opts.checkpoint)
trainer.gen_a.load_state_dict(state_dict['a'])
trainer.gen_b.load_state_dict(state_dict['b'])
except:
state_dict = pytorch03_to_pytorch04(torch.load(opts.checkpoint), opts.trainer)
trainer.gen_a.load_state_dict(state_dict['a'])
trainer.gen_b.load_state_dict(state_dict['b'])
trainer.cuda()
trainer.eval()
encode = trainer.gen_a.encode if opts.a2b else trainer.gen_b.encode # encode function
style_encode = trainer.gen_b.encode if opts.a2b else trainer.gen_a.encode # encode function
decode = trainer.gen_b.decode if opts.a2b else trainer.gen_a.decode # decode function
if 'new_size' in config:
new_size = config['new_size']
else:
if opts.a2b==1:
new_size = config['new_size_a']
else:
trainer = UNIT_Trainer(config)
else:
sys.exit("Only support MUNIT|UNIT")
############## configure checkpoint from output_folder
checkpoint_path = find_latest_model_file(os.path.join(opts.output_folder,
'checkpoints'),
opts.checkpoint,
keyword='gen')
try:
state_dict = torch.load(checkpoint_path)
trainer.gen_a.load_state_dict(state_dict['a'])
trainer.gen_b.load_state_dict(state_dict['b'])
except:
state_dict = pytorch03_to_pytorch04(torch.load(checkpoint_path))
trainer.gen_a.load_state_dict(state_dict['a'])
trainer.gen_b.load_state_dict(state_dict['b'])
trainer.cuda()
trainer.eval()
##### Dataset building
Dataset = choose_dataset(config['dataset_name'])
dataset = Dataset(config['data_root'], config, split='val')
dataloader = torch.utils.data.DataLoader(dataset,
batch_size=1,
drop_last=True,
shuffle=False,
sys.exit("Only support AttnMUNIT|MUNIT|UNIT")
# if opts.trainer == 'MUNIT':
# trainer = MUNIT_Trainer(config)
# elif opts.trainer == 'UNIT':
# trainer = UNIT_Trainer(config)
# elif opts.trainer == 'AttnMUNIT':
# trainer = AttnMUNIT_Trainer(config)
# else:
# sys.exit("Only support AttnMUNIT|MUNIT|UNIT")
try:
state_dict = torch.load(opts.checkpoints[count])
trainer.gen_a.load_state_dict(state_dict['a'])
trainer.gen_b.load_state_dict(state_dict['b'])
except:
state_dict = pytorch03_to_pytorch04(torch.load(opts.checkpoints[count]), trainer_name)
trainer.gen_a.load_state_dict(state_dict['a'])
trainer.gen_b.load_state_dict(state_dict['b'])
trainer.cuda()
trainer.eval()
encode = None
style_encode = None
decode = None
style_decode = None
if trainer_name in ['AttnMUNIT','MUNIT']:
encode = trainer.gen_a.encode if opts.a2b else trainer.gen_b.encode # encode function
style_encode = trainer.gen_b.encode if opts.a2b else trainer.gen_a.encode # encode function
decode = trainer.gen_b.decode if opts.a2b else trainer.gen_a.decode # decode function
style_decode = trainer.gen_a.decode if opts.a2b else trainer.gen_b.decode # decode function
config = get_config(opts.config)
config['vgg_model_path'] = opts.output_path
if not os.path.exists(opts.output_folder):
os.makedirs(opts.output_folder)
trainer = MUNIT_Trainer(config, device)
try:
state_dict = torch.load(opts.checkpoint)
trainer.dis_a.load_state_dict(state_dict['a'])
trainer.dis_b.load_state_dict(state_dict['b'])
except:
state_dict = pytorch03_to_pytorch04(torch.load(opts.checkpoint), 'MUNIT')
trainer.dis_a.load_state_dict(state_dict['a'])
trainer.dis_b.load_state_dict(state_dict['b'])
dis = trainer.dis_b if opts.a2b else trainer.dis_a
trainer.to(device)
trainer.eval()
transform_list = [transforms.Resize(config['crop_image_width'])]
if opts.centercrop:
transform_list.append(transforms.CenterCrop((config['crop_image_height'], config['crop_image_width'])))
transform_list.extend([transforms.ToTensor(), transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])
transform = transforms.Compose(transform_list)
input_list = [ [],[],[] ]
config['vgg_model_path'] = opts.output_path
if opts.trainer == 'MUNIT':
style_dim = config['gen']['style_dim']
trainer = MUNIT_Trainer(config, device)
elif opts.trainer == 'UNIT':
trainer = UNIT_Trainer(config, device)
else:
sys.exit("Only support MUNIT|UNIT")
try:
state_dict = torch.load(opts.checkpoint)
trainer.gen_a.load_state_dict(state_dict['a'])
trainer.gen_b.load_state_dict(state_dict['b'])
except:
state_dict = pytorch03_to_pytorch04(
torch.load(opts.checkpoint, map_location=lambda storage, loc: storage),
opts.trainer)
trainer.gen_a.load_state_dict(state_dict['a'])
trainer.gen_b.load_state_dict(state_dict['b'])
trainer.to(device=device)
trainer.eval()
encode = trainer.gen_a.encode if opts.a2b else trainer.gen_b.encode # encode function
style_encode = trainer.gen_b.encode if opts.a2b else trainer.gen_a.encode # encode function
decode = trainer.gen_b.decode if opts.a2b else trainer.gen_a.decode # decode function
if 'new_size' in config:
new_size = config['new_size']
else:
if opts.a2b == 1:
new_size = config['new_size_a']