Pytorch學習（七）---- 儲存提取

莫煩python視訊學習筆記 視訊連結https://www.bilibili.com/video/BV1Vx411j7kT?from=search&seid=3065687802317837578

import torch
from torch.autograd import Variable
import matplotlib.pyplot as plt
import os
os.environ["KMP_DUPLICATE_LIB_OK"] = "TRUE"
# import os 這裡是為了防止報錯加的

torch.manual_seed(1)         # reproducible

# fake data
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)

# 神經網路的儲存
def save():
    # save net1
    net1 = torch.nn.Sequential(              # Sequential的功能在這個括號裡逐層搭建神經層
        torch.nn.Linear(1, 10),
        torch.nn.ReLU(),                     # 激勵函式
        torch.nn.Linear(10, 1)
    )
    optimizer = torch.optim.SGD(net1.parameters(), lr=0.25)  # 傳入引數，lr是學習效率
    loss_func = torch.nn.MSELoss()

    for t in range(100):
        prediction = net1(x)
        loss = loss_func(prediction, y)  # 預測值與真實值對比
        optimizer.zero_grad()  # 將梯度降為零
        loss.backward()
        optimizer.step()  # 以學習效率0.5優化梯度

    # plot result
    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)

    torch.save(net1, 'net.pkl')  # entire all
    torch.save(net1.state_dict(), 'net_params.pkl')  # 儲存引數


# 神經網路的提取
# method 1
def restore_net():
    net2 = torch.load('net.pkl')
    prediction = net2(x)
    # plot result
    plt.figure(1, figsize=(10, 3))
    plt.subplot(132)
    plt.title('Net2')
    plt.scatter(x.data.numpy(), y.data.numpy())
    plt.plot(x.data.numpy(), prediction.data.numpy(), 'r-', lw=5)


# method 2
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_params.pkl'))
    prediction = net3(x)
    # plot result
    plt.figure(1, figsize=(10, 3))
    plt.subplot(133)
    plt.title('Net3')
    plt.scatter(x.data.numpy(), y.data.numpy())
    plt.plot(x.data.numpy(), prediction.data.numpy(), 'r-', lw=5)
    plt.show()

# save net1
save()
# restore entire net
restore_net()
# restore params
restore_params()

一開始將訓練的學習效率設定為0.5，即：

 optimizer = torch.optim.SGD(net1.parameters(), lr=0.5)  # 傳入引數，lr是學習效率

結果影像如下，很明顯學習效果是失敗的。

重新設定學習效率，發現在0.25~0.3學習效果最好