dataloader_val = torch.utils.data.DataLoader(dataset=dataset_val, batch_size=batch_size, num_workers=10, collate_fn=collate, shuffle=False, pin_memory=True)
dataset_test = IDRID(data_path, val_df, label=False)
dataloader_test = torch.utils.data.DataLoader(dataset=dataset_test, batch_size=batch_size, num_workers=10, collate_fn=collate_test, shuffle=False, pin_memory=True)
##### sizes are (w, h) ##############################
# make sure margin / 32 is over 1.5 AND size_g is divisible by 4
size_g = (args.size_g, args.size_g) # resized global image
size_p = (args.size_p, args.size_p) # cropped local patch size
sub_batch_size = args.sub_batch_size # batch size for train local patches
###################################
print("creating models......")
path_g = os.path.join(model_path, args.path_g)
path_g2l = os.path.join(model_path, args.path_g2l)
path_l2g = os.path.join(model_path, args.path_l2g)
model, global_fixed = create_model_load_weights(n_class, mode, evaluation, path_g=path_g, path_g2l=path_g2l, path_l2g=path_l2g)
###################################
num_epochs = args.num_epochs
learning_rate = args.lr
lamb_fmreg = args.lamb_fmreg
optimizer = get_optimizer(model, mode, learning_rate=learning_rate)
scheduler = LR_Scheduler('poly', learning_rate, num_epochs, len(dataloader_train))
##################################
criterion1 = FocalLoss(gamma=3)
criterion2 = nn.CrossEntropyLoss()
criterion3 = lovasz_softmax
criterion = lambda x,y: criterion1(x, y)
print("preparing datasets and dataloaders......")
batch_size = args.batch_size
data_time = AverageMeter("DataTime", ':6.3f')
batch_time = AverageMeter("BatchTime", ':6.3f')
dataset_train = get_dataset(data_path, meta_path, train=True)
dataloader_train = DataLoader(dataset_train, num_workers=4, batch_size=batch_size, collate_fn=collate, shuffle=True, pin_memory=True)
dataset_val = get_dataset(data_path, meta_path, train=False)
dataloader_val = DataLoader(dataset_val, num_workers=4, batch_size=batch_size, collate_fn=collate, shuffle=False, pin_memory=True)
###################################
print("creating models......")
path = os.path.join(model_path, args.path_test) if args.path_test else args.path_test
model = create_model_load_weights(n_class, evaluation, path=path)
###################################
num_epochs = args.epoch
learning_rate = args.lr
momentum = args.momentum
weight_decay = args.weight_decay
opt_args = dict(lr=learning_rate, momentum=momentum, weight_decay=weight_decay)
optimizer = get_optimizer(model, **opt_args)
scheduler = LR_Scheduler('poly', learning_rate, num_epochs, len(dataloader_train))
##################################
criterion = BCELoss()
if not evaluation:
writer = SummaryWriter(log_dir=log_path + task_name)