num_input = data.shape[1] - 1
full_input = data[:, 0:num_input]
full_target = data[:, num_input:num_input + 1]
train_dataset = torch.utils.data.TensorDataset(full_input, full_target)
train_loader = torch.utils.data.DataLoader(train_dataset, batch_size=97)
# choose network architecture
if args.net == 'polar':
net = PolarNet(args.hid)
else:
net = RawNet(args.hid)
if list(net.parameters()):
# initialize weight values
for m in list(net.parameters()):
m.data.normal_(0, args.init)
# use Adam optimizer
optimizer = torch.optim.Adam(net.parameters(),
eps=0.000001,
lr=args.lr,
betas=(0.9, 0.999),
weight_decay=0.0001)
# training loop
for epoch in range(1, args.epochs):
accuracy = train(net, train_loader, optimizer)
if epoch % 100 == 0 and accuracy == 100:
full_input = data[:,0:num_input]
full_target = data[:,num_input:num_input+1]
train_dataset = torch.utils.data.TensorDataset(full_input,full_target)
train_loader = torch.utils.data.DataLoader(train_dataset,batch_size=97)
# create neural network
if args.net == 'polar':
net = PolarNet(args.hid)
elif args.net == 'short':
net = ShortNet(args.hid)
else:
net = RawNet(args.hid)
if list(net.parameters()):
# initialize weight values
for m in list(net.parameters()):
m.data.normal_(0,args.init)
optimizer = torch.optim.Adam(net.parameters(),eps=0.000001,lr=args.lr,
betas=(0.9,0.999),weight_decay=0.0001)
for epoch in range(1, args.epochs):
accuracy = train(net, train_loader, optimizer)
if epoch % 100 == 0 and accuracy == 100:
break
# save model
for layer in [1,2]: