F1 = Predictor_deep_attributes(num_class=len(class_list),inc=inc)
else:
F1 = Predictor_attributes(num_class=len(class_list), inc=inc, temp=args.T)
"""
# Loading the model weights from the checkpoint
filename = "save_model_ssda/ours.ckpt.best.pth.tar"
main_dict = torch.load(filename)
args.step = main_dict['step']
print("Inferencing is being done with model at step: ", args.step)
print("best accuracy, ", main_dict['best_acc_test'])
print(filename)
G.cuda()
F1.cuda()
G.load_state_dict(main_dict['G_state_dict'])
F1.load_state_dict(main_dict['F1_state_dict'])
im_data_t = torch.FloatTensor(1)
gt_labels_t = torch.LongTensor(1)
im_data_t = im_data_t.cuda()
gt_labels_t = gt_labels_t.cuda()
im_data_t = Variable(im_data_t)
gt_labels_t = Variable(gt_labels_t)
if os.path.exists(args.checkpath) == False:
os.mkdir(args.checkpath)
"""
def eval(loader, output_file="output.txt"):
G.eval()
if "resnet" in args.net:
F1 = Predictor_deep(num_class=len(class_list),
inc=inc)
else:
F1 = Predictor(num_class=len(class_list), inc=inc, cosine=True, temp=args.T)
G.cuda()
F1.cuda()
G.load_state_dict(torch.load(os.path.join(args.checkpath,
"G_iter_model_{}_{}_"
"to_{}_step_{}.pth.tar".
format(args.method, args.source,
args.target, args.step))))
F1.load_state_dict(torch.load(os.path.join(args.checkpath,
"F1_iter_model_{}_{}_"
"to_{}_step_{}.pth.tar".
format(args.method, args.source,
args.target, args.step))))
im_data_t = torch.FloatTensor(1)
gt_labels_t = torch.LongTensor(1)
im_data_t = im_data_t.cuda()
gt_labels_t = gt_labels_t.cuda()
im_data_t = Variable(im_data_t)
gt_labels_t = Variable(gt_labels_t)
if os.path.exists(args.checkpath) == False:
os.mkdir(args.checkpath)
