# loss and network
criterion = nn.CrossEntropyLoss()
net = ResNet14(in_channels=3, out_channels=5, D=2)
print(net)
# a data loader must return a tuple of coords, features, and labels.
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
net = net.to(device)
optimizer = SGD(net.parameters(), lr=1e-2)
for i in range(10):
optimizer.zero_grad()
# Get new data
coords, feat, label = data_loader()
input = ME.SparseTensor(feat, coords=coords).to(device)
label = label.to(device)
# Forward
output = net(input)
# Loss
loss = criterion(output.F, label)
print('Iteration: ', i, ', Loss: ', loss.item())
# Gradient
loss.backward()
optimizer.step()
# Saving and loading a network
# loss and network
criterion = nn.CrossEntropyLoss()
net = MinkUNet14A(in_channels=3, out_channels=5, D=2)
print(net)
# a data loader must return a tuple of coords, features, and labels.
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
net = net.to(device)
optimizer = SGD(net.parameters(), lr=1e-2)
for i in range(10):
optimizer.zero_grad()
# Get new data
coords, feat, label = data_loader(is_classification=False)
input = ME.SparseTensor(feat, coords=coords).to(device)
label = label.to(device)
# Forward
output = net(input)
# Loss
loss = criterion(output.F, label)
print('Iteration: ', i, ', Loss: ', loss.item())
# Gradient
loss.backward()
optimizer.step()
# Saving and loading a network