def run_train():
timestamp = time.strftime("%Y%m%d") + '-' + time.strftime("%H%M%S")
path_data_train = '../../MURA_trainval_keras'
path_data_valid = '../../MURA_valid1_keras'
path_log = '../../trained_models/' + timestamp + '-dualpathnet131-adam-augslight/tb'
model_name = 'DUALPATHNET131'
model_pretrained = True
batch_size = 16
epoch_num = 100
path_model = '../../trained_models/' + timestamp + '-dualpathnet131-adam-augslight/m-' + timestamp + '.pth.tar'
device = None
opts, _ = getopt.getopt(sys.argv[1:], "d:", ["device="])
for opt, arg in opts:
if opt in ("-d", "--device") and torch.cuda.is_available():
device = torch.device("cuda:" + str(arg))
if device is None:
print("GPU not found! Using CPU!")
device = torch.device("cpu")
print('Training NN architecture = ', model_name)
train.train(path_data_train=path_data_train,
path_data_valid=path_data_valid,
path_log=path_log,
path_model=path_model,
model_name=model_name,
model_pretrained=model_pretrained,
batch_size=batch_size,
epoch_num=epoch_num,
checkpoint=None,
device=device,
transform_train=data_augmentation.augment_transform_slight(
target_mean=np.array([124 / 255, 117 / 255, 104 / 255]),
target_std=np.array([1 / (.0167 * 255)] * 3)),
transform_valid=data_augmentation.valid_transform(
target_mean=np.array([124 / 255, 117 / 255, 104 / 255]),
target_std=np.array([1 / (.0167 * 255)] * 3)),
optimizer_fn=optimizers.adam_optimizers)
print('Testing the trained model')
train.test(path_data=path_data_valid,
path_model=path_model,
model_name=model_name,
model_pretrained=model_pretrained,
batch_size=batch_size,
device=device,
transform=data_augmentation.valid_transform(
target_mean=np.array([124 / 255, 117 / 255, 104 / 255]),
target_std=np.array([1 / (.0167 * 255)] * 3)))
def run_train():
timestamp = time.strftime("%Y%m%d") + '-' + time.strftime("%H%M%S")
path_data_train = '../../MURA_trainval_keras'
path_data_valid = '../../MURA_valid1_keras'
path_log = '../../trained_models/' + timestamp + '-seresnext50_32x4d-adam-augslight-revised/tb'
model_name = 'SERESNEXT50_32x4d'
model_pretrained = True
batch_size = 16
epoch_num = 100
path_model = '../../trained_models/' + timestamp + '-seresnext50_32x4d-adam-augslight-revised/m-' + timestamp + '.pth.tar'
device = None
opts, _ = getopt.getopt(sys.argv[1:], "d:", ["device="])
for opt, arg in opts:
if opt in ("-d", "--device") and torch.cuda.is_available():
device = torch.device("cuda:" + str(arg))
if device is None:
print("GPU not found! Using CPU!")
device = torch.device("cpu")
print('Training NN architecture = ', model_name)
train.train(
path_data_train=path_data_train,
path_data_valid=path_data_valid,
path_log=path_log,
path_model=path_model,
model_name=model_name,
model_pretrained=model_pretrained,
batch_size=batch_size,
epoch_num=epoch_num,
checkpoint=None,
device=device,
transform_train=data_augmentation.augment_transform_slight_revised(
target_mean=np.array([0.485, 0.456, 0.406]),
target_std=np.array([0.229, 0.224, 0.225])),
transform_valid=data_augmentation.valid_transform(
target_mean=np.array([0.485, 0.456, 0.406]),
target_std=np.array([0.229, 0.224, 0.225])),
optimizer_fn=optimizers.adam_optimizers)
print('Testing the trained model')
train.test(path_data=path_data_valid,
path_model=path_model,
model_name=model_name,
model_pretrained=model_pretrained,
batch_size=batch_size,
device=device,
transform=data_augmentation.valid_transform())