對比各種優化器的效果
數據分佈如下圖:
import torch
import torch.utils.data as Data
import torch.nn.functional as Func
from matplotlib import pyplot as plt
torch.manual_seed(1)
#hyper parameters
LR=0.01
BATCH_SIZE=32
EPOCH=12
#fake dataset
x=torch.unsqueeze(torch.linspace(-1,1,1000),dim=1)
y=x.pow(2)+0.1*torch.normal(torch.zeros(x.size()))#torch.normal(means,std,out)
#plot dataset
# plt.scatter(x.numpy(),y.numpy())
# plt.show()
#用DataLoader來封裝數據,轉換爲Dataset形式
torch_dataset=Data.TensorDataset(x,y)
loader=Data.DataLoader(dataset=torch_dataset,batch_size=BATCH_SIZE,shuffle=True)
#network
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):
next_x=Func.relu(self.hidden(x))
pred=self.predict(next_x)
return pred
#爲每個優化器創建一個NET
net_SGD =Net()
net_Momentum =Net()
net_RMSprop =Net()
net_Adam =Net()
nets=[net_SGD,net_Momentum,net_RMSprop,net_Adam]
#創建優化器
opt_SGD =torch.optim.SGD(net_SGD.parameters(),lr=LR)
opt_Momentum=torch.optim.SGD(net_Momentum.parameters(),lr=LR,momentum=0.8)
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]
loss_func=torch.nn.MSELoss()
losses_history=[[],[],[],[]] #記錄訓練時不同神經網絡的loss
#training
for epoch in range(EPOCH):
for step,(batch_x,batch_y) in enumerate(loader):
for net,opt,l_history in zip(nets,optimizers,losses_history):
predict=net(batch_x)
loss=loss_func(predict,batch_y)
opt.zero_grad()
loss.backward()
opt.step()
l_history.append(loss.data.numpy())
labels=['SGD','Momentum','RMSprop','Adam']
for i,l_his in enumerate(losses_history):
plt.plot(l_his,label=labels[i],lw=1)
plt.legend(loc='best') #標籤放置位置
plt.xlabel('Steps')
plt.ylabel('Loss')
plt.ylim(0,0.2)
plt.show()
結果圖: