1. 官网描述
使用 torch.nn 模块，我们可以创建自定义的神经网络模型，并使用 PyTorch 提供的优化器（如 torch.optim）和损失函数来训练和优化模型。
2. 常见结构
1.卷积操作
定义：
二维卷积
1.1版本 nn.functional.conv2d
python
torch.nn.functional.conv2d(input,
weight,
bias=None,
stride=1,
padding=0,
dilation=1, groups=1)例子：
代码：
python
import torch
import torch.nn.functional as F
input = torch.tensor([[1,2,0,3,1],
[0,1,2,3,1],
[1,2,1,0,0],
[5,2,3,1,1],
[2,1,0,1,1]])
kernel = torch.tensor([[1,2,1],
[0,1,0],
[2,1,0]])
print(input.shape,kernel.shape)#5,5 3,3
input = torch.reshape(input,(1,1,5,5))
kernel = torch.reshape(kernel,(1,1,3,3))
output = F.conv2d(input,kernel,stride=1)
print(output)
output2 = F.conv2d(input,kernel,stride=2)
print(output2)
# padding=1 上下左右加一圈0
output3 = F.conv2d(input,kernel,stride=1,padding=1)
print(output3)运行：
1.2版本 torch.nn.Conv2d
python
torch.nn.Conv2d(
in_channels, 输入通道数
out_channels, 输出通道数
kernel_size, 卷积核大小（int or turple）3 (1,2)
stride=1, 卷积核移动的步长
padding=0, 边界填充（加几圈0）
dilation=1,groups=1, bias=True, padding_mode='zeros', device=None, dtype=None)代码
python
import torch
import torchvision
from torch import nn
from torch.utils.data import DataLoader
from torch.utils.tensorboard import SummaryWriter
dataset = torchvision.datasets.CIFAR10('data',train=False,
transform=torchvision.transforms.ToTensor(),
download=True)
dataloader = DataLoader(dataset,batch_size=64)
class Tudui(nn.Module):
def __init__(self):
super(Tudui,self).__init__()
self.conv1 = nn.Conv2d(3,6,3,stride=1,padding=0)
def forward(self,x):
x = self.conv1(x)
return x
tudui = Tudui()
writer = SummaryWriter('logs')
step = 0
for data in dataloader:
imgs,targets = data
output = tudui(imgs)
print(output.shape)
# 输入大小
writer.add_images('input',imgs,step)
# 输出大小
output = torch.reshape(output,(-1,3,30,30))
writer.add_images('output', output, step)
step+=1输出：
2.池化操作
目的：保留输入特征，减小数据量
最大池化MaxPool2d：下采样
torch.nn.MaxPool2d(
kernel_size, int/tuple 窗口
stride=None, 步长（默认值是Kernel_size）
padding=0, 在外面加一圈
dilation=1, （空洞卷积）
return_indices=False,
ceil_mode=False ceil模式：向上取整
)
例子：
python
import torch
import torchvision.datasets
from torch import nn
input = torch.tensor([[1,2,0,3,1],
[0,1,2,3,1],
[1,2,1,0,0],
[5,2,3,1,1],
[2,1,0,1,1]])
# (N,C,Hin,Win) (C,Hin,Win) -1表示自己计算batchsize,
input = torch.reshape(input,(-1,1,5,5))
print(input.shape)
class Tudui(nn.Module):
def __init__(self):
super(Tudui,self).__init__()
self.maxpool1 = nn.MaxPool2d(kernel_size=3,ceil_mode=True)# 模板创建自己的
def forward(self,input):
output = self.maxpool1(input)#输入
return output
tudui = Tudui()
output = tudui(input)
print(output)输出
例子2
代码
python
import torch
import torchvision.datasets
from torch import nn
from torch.utils.data import DataLoader
from torch.utils.tensorboard import SummaryWriter
class Tudui(nn.Module):
def __init__(self):
super(Tudui,self).__init__()
self.maxpool1 = nn.MaxPool2d(kernel_size=3,ceil_mode=True)# 模板创建自己的
def forward(self,input):
output = self.maxpool1(input)#输入
return output
tudui = Tudui()
dataset = torchvision.datasets.CIFAR10('data',train=False,download=True,transform=torchvision.transforms.ToTensor())
dataloader = DataLoader(dataset,batch_size=64)
writer = SummaryWriter('logs_maxpool')
step = 0
for data in dataloader:
imgs,targets = data
writer.add_images('input', imgs, step)
output = tudui(imgs)
writer.add_images('maxpool',output,step)
step+=1
writer.close()输出
3.非线性激活
python
torch.nn.ReLU(inplace=False)
torch.nn.Sigmoid(*args, **kwargs)代码：
python
import torch
from torch import nn
input = torch.tensor([[1,-0.5],
[-1,3]])
input = torch.reshape(input,(-1,1,2,2))
class Tudui(nn.Module):
def __init__(self):
super().__init__()
self.relu1 = nn.ReLU()
def forward(self,input):
output = self.relu1(input)
return output
tudui = Tudui()
output = tudui(input)
print(output)运行
例子2：
代码：
python
import torch
import torchvision.datasets
from torch import nn
from torch.utils.data import DataLoader
from torch.utils.tensorboard import SummaryWriter
# sigmoid
dataset = torchvision.datasets.CIFAR10('data',train=False,download=True,transform=torchvision.transforms.ToTensor())
dataloader = DataLoader(dataset,batch_size=64)
class Tudui(nn.Module):
def __init__(self):
super().__init__()
self.sigmoid1 = nn.Sigmoid()
def forward(self,input):
output = self.sigmoid1(input)
return output
tudui = Tudui()
writer = SummaryWriter('logs_sigmoid')
step = 0
for data in dataloader:
imgs,targets = data
writer.add_images('input',imgs,step)
output = tudui(imgs)
writer.add_images('output', output, step)
print(output.shape)
step+=1
writer.close()输出
4.线性层
torch.nn.Linear(
in_features,
out_features,
bias=True,
device=None, dtype=None)
代码：
python
import torch
import torchvision
from torch import nn
from torch.utils.data import DataLoader
dataset = torchvision.datasets.CIFAR10('data',train=False,transform=torchvision.transforms.ToTensor(),download=True)
dataloader = DataLoader(dataset,64,drop_last=True)
class Tudui(nn.Module):
def __init__(self):
super(Tudui,self).__init__()
self.linear1 = nn.Linear(in_features=196608,out_features=10)
def forward(self,input):
return self.linear1(input)
tudui = Tudui()
for data in dataloader:
imgs, targets = data
# output = torch.flatten(imgs) #torch.Size([196608])
# 替换展平
output = torch.reshape(imgs,(1,1,1,-1)) #torch.Size([1, 1, 1, 196608])
print(output.shape)
output = tudui(output) #torch.Size([10]) torch.Size([1, 1, 1, 10])
print(output.shape)4.序列化层（sequential）
例子：
代码：
python
import torch
from torch import nn
from torch.utils.tensorboard import SummaryWriter
class Tudui(nn.Module):
def __init__(self):
super().__init__()
# 草稿中计算 stride padding
self.conv1 = nn.Conv2d(in_channels=3,out_channels=32,kernel_size=5,stride=1,padding=2)
self.maxpool1 = nn.MaxPool2d(2)
self.conv2 = nn.Conv2d(in_channels=32,out_channels=32,kernel_size=5,padding=2)
self.maxpool2 = nn.MaxPool2d(2)
self.conv3 = nn.Conv2d(in_channels=32,out_channels=64,kernel_size=5,padding=2)
self.maxpool3 = nn.MaxPool2d(2)
self.flatten = nn.Flatten()
#64是隐藏单元数
self.linear1 = nn.Linear(in_features=1024,out_features=64)
#10是输出类别
self.linear2 = nn.Linear(in_features=64, out_features=10)
# 另一种写法
self.model1 = nn.Sequential(
nn.Conv2d(in_channels=3, out_channels=32, kernel_size=5, stride=1, padding=2),
nn.MaxPool2d(2),
nn.Conv2d(in_channels=32, out_channels=32, kernel_size=5, stride=1, padding=2),
nn.MaxPool2d(2),
nn.Conv2d(in_channels=32, out_channels=64, kernel_size=5, stride=1, padding=2),
nn.MaxPool2d(2),
nn.Flatten(),
nn.Linear(in_features=1024, out_features=64),
nn.Linear(in_features=64, out_features=10)
)
def forward(self,x):
# x = self.conv1(x)
# x = self.maxpool1(x)
# x = self.conv2(x)
# x = self.maxpool2(x)
# x = self.conv3(x)
# x = self.maxpool3(x)
# x = self.flatten(x)
# x = self.linear1(x)
# x = self.linear2(x)
# sequential方式
x = self.model1(x)
return x
tudui = Tudui()
print(tudui)
# 用来检验网络结构参数
input = torch.ones((64,3,32,32))# 64 batchsize（64张）
output = tudui(input)
writer= SummaryWriter('logs_s')
writer.add_graph(tudui,input)
writer.close()输出
