def sample_architecture_from_the_supernet(unique_name_of_arch, hardsampling=True):
logger = get_logger(CONFIG_SUPERNET['logging']['path_to_log_file'])
lookup_table = LookUpTable()
model = FBNet_Stochastic_SuperNet(lookup_table, cnt_classes=10).cuda()
model = nn.DataParallel(model)
load(model, CONFIG_SUPERNET['train_settings']['path_to_save_model'])
ops_names = [op_name for op_name in lookup_table.lookup_table_operations]
cnt_ops = len(ops_names)
arch_operations=[]
if hardsampling:
for layer in model.module.stages_to_search:
arch_operations.append(ops_names[np.argmax(layer.thetas.detach().cpu().numpy())])
else:
rng = np.linspace(0, cnt_ops - 1, cnt_ops, dtype=int)
for layer in model.module.stages_to_search:
distribution = softmax(layer.thetas.detach().cpu().numpy())
arch_operations.append(ops_names[np.random.choice(rng, p=distribution)])
logger.info("Sampled Architecture: " + " - ".join(arch_operations))
writh_new_ARCH_to_fbnet_modeldef(arch_operations, my_unique_name_for_ARCH=unique_name_of_arch)
logger.info("CONGRATULATIONS! New architecture " + unique_name_of_arch \
+ " was written into fbnet_building_blocks/fbnet_modeldef.py")
def main():
manual_seed = 1
np.random.seed(manual_seed)
torch.manual_seed(manual_seed)
torch.cuda.manual_seed_all(manual_seed)
torch.backends.cudnn.benchmark = True
create_directories_from_list(
[CONFIG_ARCH['logging']['path_to_tensorboard_logs']])
logger = get_logger(CONFIG_ARCH['logging']['path_to_log_file'])
writer = SummaryWriter(
log_dir=CONFIG_ARCH['logging']['path_to_tensorboard_logs'])
#### DataLoading
train_loader = get_loaders(1.0, CONFIG_ARCH['dataloading']['batch_size'],
CONFIG_ARCH['dataloading']['path_to_save_data'],
logger)
valid_loader = get_test_loader(
CONFIG_ARCH['dataloading']['batch_size'],
CONFIG_ARCH['dataloading']['path_to_save_data'])
#### Model
arch = args.architecture_name
model = fbnet_builder.get_model(arch, cnt_classes=10).cuda()
model = model.apply(weights_init)
model = nn.DataParallel(model, [0])
print(model)
#### Loss and Optimizer
optimizer = torch.optim.SGD(
filter(lambda p: p.requires_grad, model.parameters()),
lr=CONFIG_ARCH['optimizer']['lr'],
momentum=CONFIG_ARCH['optimizer']['momentum'],
weight_decay=CONFIG_ARCH['optimizer']['weight_decay'])
criterion = nn.CrossEntropyLoss().cuda()
#### Scheduler
if CONFIG_ARCH['train_settings']['scheduler'] == 'MultiStepLR':
scheduler = torch.optim.lr_scheduler.MultiStepLR(
optimizer,
milestones=CONFIG_ARCH['train_settings']['milestones'],
gamma=CONFIG_ARCH['train_settings']['lr_decay'])
elif CONFIG_ARCH['train_settings']['scheduler'] == 'CosineAnnealingLR':
scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(
optimizer,
T_max=CONFIG_ARCH['train_settings']['cnt_epochs'],
eta_min=0.001,
last_epoch=-1)
else:
logger.info(
"Please, specify scheduler in architecture_functions/config_for_arch"
)
#### Training Loop
trainer = TrainerArch(criterion, optimizer, scheduler, logger, writer)
trainer.train_loop(train_loader, valid_loader, model)
def train_supernet():
manual_seed = 1
np.random.seed(manual_seed)
torch.manual_seed(manual_seed)
torch.cuda.manual_seed_all(manual_seed)
torch.backends.cudnn.benchmark = True
create_directories_from_list([CONFIG_SUPERNET['logging']['path_to_tensorboard_logs']])
logger = get_logger(CONFIG_SUPERNET['logging']['path_to_log_file'])
writer = SummaryWriter(log_dir=CONFIG_SUPERNET['logging']['path_to_tensorboard_logs'])
#### LookUp table consists all information about layers
lookup_table = LookUpTable(calulate_latency=CONFIG_SUPERNET['lookup_table']['create_from_scratch'])
#### DataLoading
train_w_loader = dataloader.create_loaders(load_random_triplets=False,
batchsize=CONFIG_SUPERNET['dataloading']['batch_size'], n_triplets=3000)
train_thetas_loader = dataloader.create_loaders(load_random_triplets=False,
batchsize=CONFIG_SUPERNET['dataloading']['batch_size'],
n_triplets=1000)
test_loader = dataloader.create_test_loaders(load_random_triplets=False,
batchsize=CONFIG_SUPERNET['dataloading']['batch_size'], n_triplets=50)
#### Model
model = FBNet_Stochastic_SuperNet(lookup_table).cuda()
model = model.apply(weights_init)
model = nn.DataParallel(model, device_ids=[0])
#### Loss, Optimizer and Scheduler
criterion = SupernetLoss().cuda()
thetas_params = [param for name, param in model.named_parameters() if 'thetas' in name]
params_except_thetas = [param for param in model.parameters() if not check_tensor_in_list(param, thetas_params)]
w_optimizer = torch.optim.SGD(params=params_except_thetas,
lr=CONFIG_SUPERNET['optimizer']['w_lr'],
momentum=CONFIG_SUPERNET['optimizer']['w_momentum'],
weight_decay=CONFIG_SUPERNET['optimizer']['w_weight_decay'])
theta_optimizer = torch.optim.Adam(params=thetas_params,
lr=CONFIG_SUPERNET['optimizer']['thetas_lr'],
weight_decay=CONFIG_SUPERNET['optimizer']['thetas_weight_decay'])
last_epoch = -1
w_scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(w_optimizer,
T_max=CONFIG_SUPERNET['train_settings']['cnt_epochs'],
last_epoch=last_epoch)
#### Training Loop
trainer = TrainerSupernet(criterion, w_optimizer, theta_optimizer, w_scheduler, logger, writer, lookup_table)
trainer.train_loop(train_w_loader, train_thetas_loader, test_loader, model)
def train_supernet():
manual_seed = 1
np.random.seed(manual_seed)
torch.manual_seed(manual_seed)
torch.cuda.manual_seed_all(manual_seed)
torch.backends.cudnn.benchmark = True
create_directories_from_list([CONFIG_SUPERNET['logging']['path_to_tensorboard_logs']])
logger = get_logger(CONFIG_SUPERNET['logging']['path_to_log_file'])
writer = SummaryWriter(log_dir=CONFIG_SUPERNET['logging']['path_to_tensorboard_logs'])
#### DataLoading
train_w_loader, train_thetas_loader = get_loaders(CONFIG_SUPERNET['dataloading']['w_share_in_train'],
CONFIG_SUPERNET['dataloading']['batch_size'],
CONFIG_SUPERNET['dataloading']['path_to_save_data'],
logger)
test_loader = get_test_loader(CONFIG_SUPERNET['dataloading']['batch_size'],
CONFIG_SUPERNET['dataloading']['path_to_save_data'])
lookup_table = LookUpTable_HIGH(calculate_latency=CONFIG_SUPERNET['lookup_table']['create_from_scratch'], prune_type=args.prune)
###MODEL
model = FBNet_Stochastic_SuperNet(lookup_table, cnt_classes=10).cuda()
model = model.apply(weights_init)
model = nn.DataParallel(model, device_ids=[0])
for m in model.modules():
if isinstance(m, nn.Conv2d) or isinstance(m, nn.Linear):
prune.remove(m, 'weight')
#### Loss, Optimizer and Scheduler
criterion = SupernetLoss().cuda()
thetas_params = [param for name, param in model.named_parameters() if 'thetas' in name]
params_except_thetas = [param for param in model.parameters() if not check_tensor_in_list(param, thetas_params)]
w_optimizer = torch.optim.SGD(params=params_except_thetas,
lr=CONFIG_SUPERNET['optimizer']['w_lr'],
momentum=CONFIG_SUPERNET['optimizer']['w_momentum'],
weight_decay=CONFIG_SUPERNET['optimizer']['w_weight_decay'])
theta_optimizer = torch.optim.Adam(params=thetas_params,
lr=CONFIG_SUPERNET['optimizer']['thetas_lr'],
weight_decay=CONFIG_SUPERNET['optimizer']['thetas_weight_decay'])
last_epoch = -1
w_scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(w_optimizer,
T_max=CONFIG_SUPERNET['train_settings']['cnt_epochs'],
last_epoch=last_epoch)
#### Training Loop
trainer = TrainerSupernet(criterion, w_optimizer, theta_optimizer, w_scheduler, logger, writer, True)
trainer.train_loop(train_w_loader, train_thetas_loader, test_loader, model)
ops_names = [op_name for op_name in lookup_table.lookup_table_operations]
'''
for layer in model.module.stages_to_search:
#layer.thetas = nn.Parameter(torch.Tensor([1.0 / 1 for i in range(1)]).cuda())
print(layer.thetas)
'''
f = open("result.txt", "w")
for i, layer in enumerate(model.module.stages_to_search):
print('Layer {}: '.format(i) + ops_names[np.argmax(layer.thetas.detach().cpu().numpy())], end=" ")
f.write('Layer {}: '.format(i) + ops_names[np.argmax(layer.thetas.detach().cpu().numpy())]+'\n')
f.close()
print()
def main():
manual_seed = 1
np.random.seed(manual_seed)
torch.manual_seed(manual_seed)
torch.cuda.manual_seed_all(manual_seed)
torch.backends.cudnn.benchmark = True
create_directories_from_list(
[CONFIG_ARCH['logging']['path_to_tensorboard_logs']])
logger = get_logger(CONFIG_ARCH['logging']['path_to_log_file'])
writer = SummaryWriter(
log_dir=CONFIG_ARCH['logging']['path_to_tensorboard_logs'])
#### DataLoading
train_loader = get_loaders(1.0, CONFIG_ARCH['dataloading']['batch_size'],
CONFIG_ARCH['dataloading']['path_to_save_data'],
logger)
valid_loader = get_test_loader(
CONFIG_ARCH['dataloading']['batch_size'],
CONFIG_ARCH['dataloading']['path_to_save_data'])
#### Model
arch = args.architecture_name
model = fbnet_builder.get_model(arch, cnt_classes=10).cuda()
#model.load_state_dict(torch.load("architecture_functions/logs/best_model.pth", map_location=torch.device('cuda')))
#checkpoint = torch.load("architecture_functions/logs/best_model.pth", map_location=torch.device('cuda'))
#state_dict = torch.load("architecture_functions/logs/best_model.pth", map_location="cuda")
state_dict = torch.load(
"/home/oza/pre-experiment/speeding/FBNet/architecture_functions/logs/best_model.pth",
map_location="cuda")
#state_dict = torch.load("/home/oza/pre-experiment/speeding/test_dist/logs/test_FBnetA0826/best.pth.tar")['state_dict']
#state_dict = torch.load("/home/oza/pre-experiment/speeding/testFB/FBNet/architecture_functions/logs/fbnet_a/best_model.pth", map_location="cuda")
#state_dict = torch.load("/home/oza/pre-experiment/speeding/testFB/distiller/examples/classifier_compression/logs/2020.08.29-035310/best.pth.tar", map_location="cuda")['state_dict']
#state_dict = torch.load("/home/oza/pre-experiment/speeding/testFB/FBNet/architecture_functions/logs/best_model.pth", map_location="cuda")
#state_dict = torch.load("/home/oza/pre-experiment/speeding/testFB/FBNet/architecture_functions/logs/fbnet_a/best_model.pth", map_location="cuda")
if "model_ema" in state_dict and state_dict["model_ema"] is not None:
state_dict = state_dict["model_ema"]
ret = {}
for name, val in state_dict.items():
if name.startswith("module."):
name = name[len("module."):]
#print(name)
ret[name] = val
# ToTensor：画像のグレースケール化（RGBの0~255を0~1の範囲に正規化）、Normalize：Z値化（RGBの平均と標準偏差を0.5で決め打ちして正規化）
#transform = transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])
CIFAR_MEAN = [0.49139968, 0.48215827, 0.44653124]
CIFAR_STD = [0.2023, 0.1994, 0.2010]
transform = transforms.Compose(
[transforms.ToTensor(),
transforms.Normalize(CIFAR_MEAN, CIFAR_STD)])
# トレーニングデータをダウンロード
trainset = torchvision.datasets.CIFAR10(root='./data',
train=True,
download=True,
transform=transform)
trainloader = torch.utils.data.DataLoader(trainset,
batch_size=4,
shuffle=True,
num_workers=2)
# テストデータをダウンロード
#testset = torchvision.datasets.CIFAR10(root='./data', train=False, download=True, transform=transform)
#testloader = torch.utils.data.DataLoader(testset, batch_size=4, shuffle=True, num_workers=2)
test_data = datasets.CIFAR10(root='./data',
train=False,
download=True,
transform=transform)
testloader = torch.utils.data.DataLoader(test_data,
batch_size=4,
shuffle=False,
num_workers=16)
model.load_state_dict(ret)
model.eval()
correct = 0
total = 0
topk = (1, )
'''
for data in testloader:
images, labels = data
images = images.to('cuda')
labels = labels.to('cuda')
outputs = model(images)
maxk = max(topk)
_, pred = outputs.topk(maxk, 1, True, True)
pred = pred.t()
# one-hot case
if labels.ndimension() > 1:
labels = labels.max(1)[1]
correct = pred.eq(labels.view(1, -1).expand_as(pred))
res = []
for k in topk:
correct_k = correct[:k].view(-1).float().sum(0)
res.append(correct_k.mul_(1.0 / 10000))
print(res)
'''
i = 0
with torch.no_grad():
for data in testloader:
images, labels = data
images = images.to('cuda')
labels = labels.to('cuda')
outputs = model(images)
_, predicted = torch.max(outputs.data, 1)
total += labels.size(0)
correct += (predicted == labels).sum().item()
#print(total)
#print(correct)
i += 1
print(i)
print(total)
print('Accuracy of the network on the 10000 test images: %d %%' %
(100 * correct / total))
quit()
model = model.apply(weights_init)
model = nn.DataParallel(model, [0])
#### Loss and Optimizer
optimizer = torch.optim.SGD(
filter(lambda p: p.requires_grad, model.parameters()),
lr=CONFIG_ARCH['optimizer']['lr'],
momentum=CONFIG_ARCH['optimizer']['momentum'],
weight_decay=CONFIG_ARCH['optimizer']['weight_decay'])
criterion = nn.CrossEntropyLoss().cuda()
#### Scheduler
if CONFIG_ARCH['train_settings']['scheduler'] == 'MultiStepLR':
scheduler = torch.optim.lr_scheduler.MultiStepLR(
optimizer,
milestones=CONFIG_ARCH['train_settings']['milestones'],
gamma=CONFIG_ARCH['train_settings']['lr_decay'])
elif CONFIG_ARCH['train_settings']['scheduler'] == 'CosineAnnealingLR':
scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(
optimizer,
T_max=CONFIG_ARCH['train_settings']['cnt_epochs'],
eta_min=0.001,
last_epoch=-1)
else:
logger.info(
"Please, specify scheduler in architecture_functions/config_for_arch"
)
#### Training Loop
trainer = TrainerArch(criterion, optimizer, scheduler, logger, writer)
trainer.train_loop(train_loader, valid_loader, model)
def train_supernet():
manual_seed = 1
np.random.seed(manual_seed)
torch.manual_seed(manual_seed)
torch.cuda.manual_seed_all(manual_seed)
torch.backends.cudnn.benchmark = True
create_directories_from_list(
[CONFIG_SUPERNET['logging']['path_to_tensorboard_logs']])
logger = get_logger(CONFIG_SUPERNET['logging']['path_to_log_file'])
writer = SummaryWriter(
log_dir=CONFIG_SUPERNET['logging']['path_to_tensorboard_logs'])
#### DataLoading
train_w_loader, train_thetas_loader = get_loaders(
CONFIG_SUPERNET['dataloading']['w_share_in_train'],
CONFIG_SUPERNET['dataloading']['batch_size'],
CONFIG_SUPERNET['dataloading']['path_to_save_data'], logger)
test_loader = get_test_loader(
CONFIG_SUPERNET['dataloading']['batch_size'],
CONFIG_SUPERNET['dataloading']['path_to_save_data'])
###TRAIN HIGH_LEVEL
lookup_table = LookUpTable_HIGH(
calulate_latency=CONFIG_SUPERNET['lookup_table']
['create_from_scratch'])
if args.high_or_low == 'high':
###MODEL
model = FBNet_Stochastic_SuperNet(lookup_table, cnt_classes=10).cuda()
model = model.apply(weights_init)
model = nn.DataParallel(model, device_ids=[0])
model.load_state_dict(
torch.load('/home/khs/data/sup_logs/cifar10/pretrained_high.pth'))
#### Loss, Optimizer and Scheduler
criterion = SupernetLoss().cuda()
for layer in model.module.stages_to_search:
layer.thetas = nn.Parameter(
torch.Tensor([1.0 / 6 for i in range(6)]).cuda())
thetas_params = [
param for name, param in model.named_parameters()
if 'thetas' in name
]
params_except_thetas = [
param for param in model.parameters()
if not check_tensor_in_list(param, thetas_params)
]
w_optimizer = torch.optim.SGD(
params=params_except_thetas,
lr=CONFIG_SUPERNET['optimizer']['w_lr'],
momentum=CONFIG_SUPERNET['optimizer']['w_momentum'],
weight_decay=CONFIG_SUPERNET['optimizer']['w_weight_decay'])
theta_optimizer = torch.optim.Adam(
params=thetas_params,
lr=CONFIG_SUPERNET['optimizer']['thetas_lr'],
weight_decay=CONFIG_SUPERNET['optimizer']['thetas_weight_decay'])
last_epoch = -1
w_scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(
w_optimizer,
T_max=CONFIG_SUPERNET['train_settings']['cnt_epochs'],
last_epoch=last_epoch)
#### Training Loop
trainer = TrainerSupernet(criterion, w_optimizer, theta_optimizer,
w_scheduler, logger, writer, True)
trainer.train_loop(train_w_loader, train_thetas_loader, test_loader,
model)
ops_names = [
op_name for op_name in lookup_table.lookup_table_operations
]
f = open('result.txt', 'w')
for i, layer in enumerate(model.module.stages_to_search):
print(ops_names[np.argmax(layer.thetas.detach().cpu().numpy())],
end=" ")
f.write('Layer {}: '.format(i) +
ops_names[np.argmax(layer.thetas.detach().cpu().numpy())])
f.close()
else:
count = 0
previous = []
index = []
act_update = []
weight_update = []
while True:
print(count, "th Iterations")
lookup_table = LookUpTable(
calulate_latency=CONFIG_SUPERNET['lookup_table']
['create_from_scratch'],
count=count,
act_update=act_update,
weight_update=weight_update)
for i in range(len(weight_update)):
weight_update[i] = 0
#if count != 0:
# lookup_table.index[0] = copy.deepcopy(index)
###MODEL
model = FBNet_Stochastic_SuperNet(lookup_table,
cnt_classes=10).cuda()
model = nn.DataParallel(model, device_ids=[0])
#if count == 0:
# model.load_state_dict(torch.load('/home/khs/data/sup_logs/cifar10/pretrained.pth'))
#else:
#model.load_state_dict(torch.load('/home/khs/data/sup_logs/cifar10/best_model.pth'))
model.load_state_dict(
torch.load('/home/khs/data/sup_logs/cifar10/best_model.pth'))
#model = model.apply(weights_init)
#### Loss, Optimizer and Scheduler
criterion = SupernetLoss().cuda()
for layer in model.module.stages_to_search:
layer.thetas = nn.Parameter(
torch.Tensor([1.0 / 3 for i in range(3)]).cuda())
thetas_params = [
param for name, param in model.named_parameters()
if 'thetas' in name
]
params_except_thetas = [
param for param in model.parameters()
if not check_tensor_in_list(param, thetas_params)
]
w_optimizer = torch.optim.SGD(
params=params_except_thetas,
lr=CONFIG_SUPERNET['optimizer']['w_lr'],
momentum=CONFIG_SUPERNET['optimizer']['w_momentum'],
weight_decay=CONFIG_SUPERNET['optimizer']['w_weight_decay'])
theta_optimizer = torch.optim.Adam(
params=thetas_params,
lr=CONFIG_SUPERNET['optimizer']['thetas_lr'],
weight_decay=CONFIG_SUPERNET['optimizer']
['thetas_weight_decay'])
last_epoch = -1
w_scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(
w_optimizer,
T_max=CONFIG_SUPERNET['train_settings']['cnt_epochs'],
last_epoch=last_epoch)
#### Training Loop
trainer = TrainerSupernet(criterion, w_optimizer, theta_optimizer,
w_scheduler, logger, writer, False)
trainer.train_loop(train_w_loader, train_thetas_loader,
test_loader, model)
del index[:]
with open('index.txt', 'w') as f:
for idx, layer in enumerate(model.module.stages_to_search):
ops = np.argmax(layer.thetas.detach().cpu().numpy())
tmp = lookup_table.index[ops][idx]
index.append(tmp)
f.write('%s\n' % tmp)
f.close()
same = 1
if count != 0:
for i in range(len(previous)):
for j in range(len(previous[i])):
if previous[i][j] not in index[i]:
same = 0
if same == 1:
break
previous = copy.deepcopy(index)
count += 1
def train_supernet():
test_input = torch.rand(1, 3, 224, 224).cuda()
manual_seed = 1
np.random.seed(manual_seed)
torch.manual_seed(manual_seed)
torch.cuda.manual_seed_all(manual_seed)
torch.backends.cudnn.benchmark = True
create_directories_from_list([CONFIG_SUPERNET['logging']['path_to_tensorboard_logs']])
logger = get_logger(CONFIG_SUPERNET['logging']['path_to_log_file'])
writer = SummaryWriter(log_dir=CONFIG_SUPERNET['logging']['path_to_tensorboard_logs'])
#### DataLoading
train_w_loader, train_thetas_loader = get_loaders(CONFIG_SUPERNET['dataloading']['w_share_in_train'],
CONFIG_SUPERNET['dataloading']['batch_size'],
CONFIG_SUPERNET['dataloading']['path_to_save_data'],
logger)
test_loader = get_test_loader(CONFIG_SUPERNET['dataloading']['batch_size'],
CONFIG_SUPERNET['dataloading']['path_to_save_data'])
###TRAIN HIGH_LEVEL
lookup_table = LookUpTable_HIGH(calulate_latency=CONFIG_SUPERNET['lookup_table']['create_from_scratch'])
###MODEL
model = FBNet_Stochastic_SuperNet(lookup_table, cnt_classes=1000)
model = model.apply(weights_init)
model = nn.DataParallel(model).cuda()
model.load_state_dict(torch.load('/home/khs/data/sup_logs/imagenet/pretrained_high.pth'))
'''
#### Loss, Optimizer and Scheduler
criterion = SupernetLoss().cuda()
for layer in model.module.stages_to_search:
layer.thetas = nn.Parameter(torch.Tensor([1.0 / 6 for i in range(6)]).cuda())
thetas_params = [param for name, param in model.named_parameters() if 'thetas' in name]
params_except_thetas = [param for param in model.parameters() if not check_tensor_in_list(param, thetas_params)]
w_optimizer = torch.optim.SGD(params=params_except_thetas,
lr=CONFIG_SUPERNET['optimizer']['w_lr'],
momentum=CONFIG_SUPERNET['optimizer']['w_momentum'],
weight_decay=CONFIG_SUPERNET['optimizer']['w_weight_decay'])
theta_optimizer = torch.optim.Adam(params=thetas_params,
lr=CONFIG_SUPERNET['optimizer']['thetas_lr'],
weight_decay=CONFIG_SUPERNET['optimizer']['thetas_weight_decay'])
last_epoch = -1
w_scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(w_optimizer,
T_max=CONFIG_SUPERNET['train_settings']['cnt_epochs'],
last_epoch=last_epoch)
#### Training Loop
trainer = TrainerSupernet(criterion, w_optimizer, theta_optimizer, w_scheduler, logger, writer, True)
trainer.train_loop(train_w_loader, train_thetas_loader, test_loader, model)
'''
model = model.eval()
model2 = mobilenet_v2().cuda()
model2 = model2.eval()
out = model(test_input, 5.0)
out2 = model2(test_input)
print(out[0].detach().cpu().numpy().shape)
print(out2.detach().cpu().numpy().shape)
'''
out = out[0].detach().cpu().numpy()
out2 = out2.detach().cpu().numpy()
if not (out == out2).all():
print(out-out2)
'''
'''
