优化器optim

优化器

建立优化器实例

所有的优化方法都继承自optim.Optimizer
包含属性：

optimizer.defaults
optimizer.param_groups
optimizer.param_groups[0].keys()

torch.optim实现了大多数优化方法，如RMSProp, Adam, SGD

优化器的使用

from torch import optim
# 新建一个优化器，指定要调整的参数和学习率
optimizer= optim.SGD(net.parameters(), lr=0.01)
# 清空梯度
optimizer.zero_grad()  # 与net.zero_grad()效果相同
# 前向传播，计算损失
output= net(input)
loss= criterion(output, target)
# 反向传播
loss.backward()
# 更新参数
optimizer.step()

优化器的比较

import torch
import torch.utils.data as Data
import torch.nn.functional as F
import matplotlib.pyplot as plt
# %matplotlib inline
"""------------------ 超参数 -----------------"""
lr= 0.01
batch_size= 32
epoches= 12

"""----------------- 生成数据 -----------------"""
# torch只能处理二维的数据
x= torch.unsqueeze(torch.linspace(-1,1, 1000), dim=1)
# 0.1* torch.normal(x.size())增加噪声
y= x.pow(2)+ 0.1*torch.normal(torch.zeros(*x.size()))

torch_dataset= Data.TensorDataset(x, y)
loader= Data.DataLoader(dataset=torch_dataset, batch_size=batch_size, shuffle=True)

"""----------------- 构建神经网络 --------------"""
class Net(torch.nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.hidden= torch.nn.Linear(1, 20)
        self.predict= torch.nn.Linear(20, 1)
    def forward(self, x):
        x= F.relu(self.hidden(x))
        x= self.predict(x)
        return x

"""----------------- 使用多种优化器 ------------"""
net_SGD= Net()
net_Momentum= Net()
net_RMSProp= Net()
net_Adam= Net()

nets= [net_SGD, net_Momentum, net_RMSProp, net_Adam]

opt_SGD= torch.optim.SGD(net_SGD.parameters(), lr=lr)
opt_Momentum= torch.optim.SGD(net_Momentum.parameters(), lr=lr, momentum=0.9)
opt_RMSProp= torch.optim.RMSprop(net_RMSProp.parameters(), lr=lr, alpha=0.9)
opt_Adam= torch.optim.Adam(net_Adam.parameters(), lr=lr, betas=(0.9,0.99))

optimizers= [opt_SGD, opt_Momentum, opt_RMSProp, opt_Adam]

"""----------------- 训练模型 -----------------"""
loss_func= torch.nn.MSELoss()
loss_his= [[], [], [], []]  # 记录损失
for epoch in range(epoches):
    for step,(batch_x,batch_y) in enumerate(loader):
        for net,opt,l_his in zip(nets, optimizers, loss_his):
            output= net(batch_x)
            loss= loss_func(output, batch_y)
            opt.zero_grad()  # 清空梯度
            loss.backward()
            opt.step()
            l_his.append(loss.data.numpy())

labels= ["SGD", "Momentum","RMSprop","Adam"]
for i,l_his in enumerate(loss_his):
    plt.plot(l_his, label=labels[i])
plt.legend(loc="best")
plt.xlabel("steps")
plt.ylabel("loss")
plt.ylim((0,0.2))
plt.show()