莫烦python--构建神经网络的三种方式

最新推荐文章于 2025-04-05 17:11:49 发布

三维视觉工作室

最新推荐文章于 2025-04-05 17:11:49 发布

阅读量294

点赞数

文章标签：深度学习神经网络

本文链接：https://blog.youkuaiyun.com/weixin_41405284/article/details/108998929

版权

一.简易搭建

直接使用net2即可

net2=torch.nn.Sequential(
        torch.nn.Linear(2,10),
        torch.nn.ReLU(),
        torch.nn.Linear(10,2)
        )

二.形参传递

class Net(torch.nn.Module):
   def __init__(self,n_features,n_hidden,n_output):
       super(Net,self).__init__() 
       self.hidden=torch.nn.Linear(n_features,n_hidden)
        self.out=torch.nn.Linear(n_hidden,n_output)
    def forward(self,x):
        x=F.relu(self.hidden(x))
        x=self.out(x)
        return x

*需要自己传入网络神经元参数

net=Net(2,10,2)

三.形参赋值的形式

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(x)
        x=self.predict(x)
        return x
net_SGD         = Net()
net_Momentum    = Net()
net_RMSprop     = Net()
net_Adam        = Net()

四.训练数据

import torch.utils.data as Data
#x为训练数据，y为标签
torch_dataset=Data.TensorDataset(x,y)
#data模块加载数据集，训练批次，乱序训练
loader=Data.DataLoader(dataset=torch_dataset,batch_size=BATCH_SIZE,shuffle=True,)

迭代训练
1.知识点补充：这里for 循环enumerate ： step表示本次迭代的索引，batch_x,batch_y为训练数据和标签

#整套数据训练三次
for epoch in range(3):
    for step,(batch_x,batch_y) in enumerate(loader): #每一步loader用BATCH——SIZE训练
         print('Epoch: ', epoch, '| Step: ', step, '| batch x: ',
              batch_x.numpy(), '| batch y: ', batch_y.numpy())

在这里插入图片描述

for循环zip并行训练，net, opt, l_his in分别对应数组nets, optimizers, losses_his中的数据

 for net, opt, l_his in zip(nets, optimizers, losses_his):
               output = net(b_x)              # get output for every net
               loss = loss_func(output, b_y)  # compute loss for every net
               opt.zero_grad()                # clear gradients for next train
               loss.backward()                # backpropagation, compute gradients
               opt.step()                     # apply gradients
               l_his.append(loss.data.numpy())     # loss recoder

参考zip学习