Pytorch学习05——快速搭建、保存提取

之前的方法

class Net(torch.nn.Module):
    def __init__(self, n_feature, n_hidden, n_output):
        super(Net, self).__init__()
        self.hidden = torch.nn.Linear(n_feature, n_hidden)
        self.predict = torch.nn.Linear(n_hidden, n_output)

    def forward(self, x):
        x = F.relu(self.hidden(x))
        x = self.predict(x)
        return x

net1 = Net(1, 10, 1)   # 这是我们用这种方式搭建的 net1
print(net1)
"""
Net (
  (hidden): Linear (1 -> 10)
  (predict): Linear (10 -> 1)
)
"""

快速搭建

net2 = torch.nn.Sequential(
    torch.nn.Linear(1, 10),
    torch.nn.ReLU(),
    torch.nn.Linear(10, 1)
)
print(net2)
"""
Sequential (
  (0): Linear (1 -> 10)
  (1): ReLU ()
  (2): Linear (10 -> 1)
)
"""

在这种方法中，Net2把激励函数一起加进去了，但是 net1 中, 激励函数实际上是在 forward() 功能中才被调用的

保存和提取

保存

torch.manual_seed(1)    # reproducible

# 假数据
x = torch.unsqueeze(torch.linspace(-1, 1, 100), dim=1)  # x data (tensor), shape=(100, 1)
y = x.pow(2) + 0.2*torch.rand(x.size())  # noisy y data (tensor), shape=(100, 1)

def save():
    # 快速建网络
    net1 = torch.nn.Sequential(
        torch.nn.Linear(1, 10),
        torch.nn.ReLU(),
        torch.nn.Linear(10, 1)
    )
    optimizer = torch.optim.SGD(net1.parameters(), lr=0.5)
    loss_func = torch.nn.MSELoss()

    # 训练
    for t in range(100):
        prediction = net1(x)
        loss = loss_func(prediction, y)
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

# 两种方式保存
torch.save(net1, 'net.pkl')  # 保存整个网络
torch.save(net1.state_dict(), 'net_params.pkl')   # 只保存网络中的参数 (速度快, 占内存少)

提取

提取整个神经网络，速度会比较慢

def restore_net():
    # restore entire net1 to net2
    net2 = torch.load('net.pkl')
    prediction = net2(x)

只提取参数

def restore_params():
    # 新建 net3
    net3 = torch.nn.Sequential(
        torch.nn.Linear(1, 10),
        torch.nn.ReLU(),
        torch.nn.Linear(10, 1)
    )

    # 将保存的参数复制到 net3
    net3.load_state_dict(torch.load('net_params.pkl'))
    prediction = net3(x)