莫烦python-pytorch-3.4保存提取

保存、读取模型

保存模型

def save():
    # save net1
    net1 = torch.nn.Sequential(
        torch.nn.Linear(1, 10),
        torch.nn.ReLU(),
        torch.nn.Linear(10, 1),
    )
    optimizer = torch.optim.SGD(net1.parameters(), lr=0.2)
    loss_func = torch.nn.MSELoss()

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

    torch.save(net1, 'net.pkl')                            # entire net
    torch.save(net1.state_dict(), 'net_param.pkl')         # parameters

保存整个网络

torch.save(net1, 'net.pkl') 

保存网络参数

torch.save(net1.state_dict(), 'net_param.pkl') 

读取整个网络

def restore_net():
    net2 = torch.load('net.pkl')

读取网络参数

先建立网络结构，后加载参数。

def restore_params():
    net3 = torch.nn.Sequential(
        torch.nn.Linear(1, 10),
        torch.nn.ReLU(),
        torch.nn.Linear(10, 1),
    )
    net3.load_state_dict(torch.load('net_param.pkl'))

完整代码

import torch
import torch.nn.functional as F
from torch.autograd import Variable
import matplotlib.pyplot as plt


# 设置随机种子，使每次得到的随机数不变
torch.manual_seed(1) # reproducible

x = torch.unsqueeze(torch.linspace(-1, 1, 100), dim=1)
y = x.pow(2) + 0.2*torch.rand(x.size())
x, y = Variable(x, requires_grad=False), Variable(y, requires_grad=False)


def save():
    # save net1
    net1 = torch.nn.Sequential(
        torch.nn.Linear(1, 10),
        torch.nn.ReLU(),
        torch.nn.Linear(10, 1),
    )
    optimizer = torch.optim.SGD(net1.parameters(), lr=0.2)
    loss_func = torch.nn.MSELoss()

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

    torch.save(net1, 'net.pkl')                            # entire net
    torch.save(net1.state_dict(), 'net_param.pkl')         # parameters

    # plot results
    plt.figure(1, figsize=(10, 3))
    plt.subplot(131)
    plt.title('Net1')
    plt.scatter(x.data.numpy(), y.data.numpy())
    plt.plot(x.data.numpy(), prediction.data.numpy(), 'r-', lw=5)


def restore_net():
    net2 = torch.load('net.pkl')
    prediction = net2(x)
    # plot results
    plt.figure(1, figsize=(10, 3))
    plt.subplot(132)
    plt.title('Net1')
    plt.scatter(x.data.numpy(), y.data.numpy())
    plt.plot(x.data.numpy(), prediction.data.numpy(), 'r-', lw=5)


def restore_params():
    net3 = torch.nn.Sequential(
        torch.nn.Linear(1, 10),
        torch.nn.ReLU(),
        torch.nn.Linear(10, 1),
    )
    net3.load_state_dict(torch.load('net_param.pkl'))
    prediction = net3(x)
    # plot results
    plt.figure(1, figsize=(10, 3))
    plt.subplot(133)
    plt.title('Net1')
    plt.scatter(x.data.numpy(), y.data.numpy())
    plt.plot(x.data.numpy(), prediction.data.numpy(), 'r-', lw=5)


save()
restore_net()
restore_params()
plt.show()