#plt.savefig('log/fed_{}_{}_{}_C{}_iid{}.png'.format(args.dataset, args.model, args.epochs, args.frac, args.iid))
#print("COUNT DATA",str(count_array))
print("NO ATTACK DATA=",loss_train)
print("1 ATTACK DATA=",loss_train_1)
print("5 ATTACK DATA=",loss_train_5)
print("10 ATTACK DATA=",loss_train_10)
print("15 ATTACK DATA=",loss_train_15)
print("20 ATTACK DATA=",loss_train_20)
print("25 ATTACK DATA=",loss_train_25)
print("30 ATTACK DATA=",loss_train_30)
print(malicious_dict)
print(malicious_count)
# testing
net_glob.eval()
#print("Agent_Found_Count",agent_found_count)
acc_train, loss_train, acc_train0,acc_train1,acc_train2,acc_train3,acc_train4,acc_train5,acc_train6,acc_train7,acc_train8,acc_train9,target_train_freq,correct_train_freq = test_img(net_glob, dataset_train, args)
acc_test, loss_test, acc_test0,acc_test1,acc_test2,acc_test3,acc_test4,acc_test5,acc_test6,acc_test7,acc_test8,acc_test9,target_test_freq,correct_test_freq = test_img(net_glob, dataset_test, args)
print("Training accuracy (NO ATTACK): {:.2f}".format(acc_train))
print("Testing accuracy (NO ATTACK): {:.2f}".format(acc_test))
print("Training accuracy-airplane (NO ATTACK): {:.2f}".format(acc_train0))
print("Testing accuracy-airplane (NO ATTACK): {:.2f}".format(acc_test0))
print("\n")
net_glob1.eval()
acc_train_1, loss_train_1, acc_train0_1,acc_train1_1,acc_train2_1,acc_train3_1,acc_train4_1,acc_train5_1,acc_train6_1,acc_train7_1,acc_train8_1,acc_train9_1,target_train_freq_1,correct_train_freq_1 = test_img(net_glob1, dataset_train, args)
acc_test_1, loss_test_1, acc_test0_1,acc_test1_1,acc_test2_1,acc_test3_1,acc_test4_1,acc_test5_1,acc_test6_1,acc_test7_1,acc_test8_1,acc_test9_1,target_test_freq_1,correct_test_freq_1 = test_img(net_glob1, dataset_test, args)
print("Training accuracy (LABEL FLIPPED 1): {:.2f}".format(acc_train_1))
print("Testing accuracy (LABEL FLIPPED 1): {:.2f}".format(acc_test_1))
print("Training accuracy-airplane (LABEL FLIPPED 1): {:.2f}".format(acc_train0_1))
print("NO ATTACK DATA=",loss_train)
print("1 ATTACK DATA=",loss_train_1)
print("5 ATTACK DATA=",loss_train_5)
print("7 ATTACK DATA=",loss_train_7)
print("10 ATTACK DATA=",loss_train_10)
# testing
net_glob.eval()
#print("Agent_Found_Count",agent_found_count)
acc_train, loss_train = test_img(net_glob, dataset_train, args)
acc_test, loss_test = test_img(net_glob, dataset_test, args)
print("Training accuracy (NO ATTACK): {:.2f}".format(acc_train))
print("Testing accuracy (NO ATTACK): {:.2f}".format(acc_test))
net_glob1.eval()
acc_train1, loss_train_1 = test_img(net_glob1, dataset_train, args)
acc_test1, loss_test_1 = test_img(net_glob1, dataset_test, args)
print("Training accuracy (CONSTANT ATTACK 1): {:.2f}".format(acc_train1))
print("Testing accuracy (CONSTANT ATTACK 1): {:.2f}".format(acc_test1))
net_glob5.eval()
acc_train5, loss_train_5 = test_img(net_glob5, dataset_train, args)
acc_test5, loss_test_5 = test_img(net_glob5, dataset_test, args)
print("Training accuracy (CONSTANT ATTACK 5): {:.2f}".format(acc_train5))
print("Testing accuracy (CONSTANT ATTACK 5): {:.2f}".format(acc_test5))
net_glob7.eval()
acc_train7, loss_train_7 = test_img(net_glob7, dataset_train, args)
print("5 ATTACK DATA=", loss_train_5)
print("10 ATTACK DATA=", loss_train_10)
print("15 ATTACK DATA=", loss_train_15)
print("20 ATTACK DATA=", loss_train_20)
print("25 ATTACK DATA=", loss_train_25)
print("30 ATTACK DATA=", loss_train_30)
# testing
net_glob.eval()
#print("Agent_Found_Count",agent_found_count)
acc_train, loss_train = test_img(net_glob, dataset_train, args)
acc_test, loss_test = test_img(net_glob, dataset_test, args)
print("Training accuracy (NO ATTACK): {:.2f}".format(acc_train))
print("Testing accuracy (NO ATTACK): {:.2f}".format(acc_test))
net_glob1.eval()
acc_train1, loss_train_1 = test_img(net_glob1, dataset_train, args)
acc_test1, loss_test_1 = test_img(net_glob1, dataset_test, args)
print("Training accuracy (CONSTANT ATTACK 1): {:.2f}".format(acc_train1))
print("Testing accuracy (CONSTANT ATTACK 1): {:.2f}".format(acc_test1))
net_glob5.eval()
acc_train5, loss_train_5 = test_img(net_glob5, dataset_train, args)
acc_test5, loss_test_5 = test_img(net_glob5, dataset_test, args)
print("Training accuracy (CONSTANT ATTACK 5): {:.2f}".format(acc_train5))
print("Testing accuracy (CONSTANT ATTACK 5): {:.2f}".format(acc_test5))
net_glob10.eval()
acc_train10, loss_train_10 = test_img(net_glob10, dataset_train, args)
acc_test10, loss_test_10 = test_img(net_glob10, dataset_test, args)
print("Training accuracy (CONSTANT ATTACK 10): {:.2f}".format(acc_train10))
class Server():
def __init__(self, args, w):
self.args = args
self.clients_update_w = []
self.clients_loss = []
self.model = CNNMnist(args=args).to(args.device)
self.model.load_state_dict(w)
def FedAvg(self):
# 1. part one
# DP mechanism
# cause we choose to add noise at client end,the fedavg should be the same as plain
if self.args.mode == 'plain' or self.args.mode == 'DP':
update_w_avg = copy.deepcopy(self.clients_update_w[0])
for k in update_w_avg.keys():
for i in range(1, len(self.clients_update_w)):
update_w_avg[k] += self.clients_update_w[i][k]
update_w_avg[k] = torch.div(update_w_avg[k],
len(self.clients_update_w))
self.model.state_dict()[k] += update_w_avg[k]
return copy.deepcopy(self.model.state_dict()), sum(
self.clients_loss) / len(self.clients_loss)
# 2. part two
# Paillier add
elif self.args.mode == 'Paillier':
update_w_avg = copy.deepcopy(self.clients_update_w[0])
for k in update_w_avg.keys():
client_num = len(self.clients_update_w)
for i in range(1, client_num):
for iter in range(len(update_w_avg[k])):
update_w_avg[k][iter] += self.clients_update_w[i][k][
iter]
for iter in range(len(update_w_avg[k])):
update_w_avg[k][iter] /= client_num
return update_w_avg, sum(self.clients_loss) / len(
self.clients_loss)
else:
exit()
def test(self, datatest):
self.model.eval()
# testing
test_loss = 0
correct = 0
data_loader = DataLoader(datatest, batch_size=self.args.bs)
for idx, (data, target) in enumerate(data_loader):
if self.args.gpu != -1:
data, target = data.cuda(), target.cuda()
log_probs = self.model(data)
# sum up batch loss
test_loss += F.cross_entropy(log_probs, target,
reduction='sum').item()
# get the index of the max log-probability
y_pred = log_probs.data.max(1, keepdim=True)[1]
correct += y_pred.eq(
target.data.view_as(y_pred)).long().cpu().sum()
test_loss /= len(data_loader.dataset)
accuracy = 100.00 * correct / len(data_loader.dataset)
return accuracy, test_loss