编程基础很弱,需要机器学习,学习记录,按自己理解写的,希望以后能学懂吧,要是有大神看到还请赐教。
神经网络-卷积层
iimport torch
import torchvision
from torch import nn
from torch.nn import Conv2d
from torch.utils.data import DataLoader
from torch.utils.tensorboard import SummaryWriter
cifar_path = "D:\\Scientific_Research\\Pycharmproject\\Biji\\dataset"
dataset = torchvision.datasets.CIFAR10(root=cifar_path, train=False, transform=torchvision.transforms.ToTensor(),
download=True)
dataloader = DataLoader(dataset=dataset, batch_size=64)
# 搭建神经网络
class Mynn(nn.Module):
# 初始化
def __init__(self):
super(Mynn, self).__init__()
# 设置卷积参数,输入图像通道数3,输出为6,卷积核为3,步数1,不padding
self.conv1 = Conv2d(in_channels=3, out_channels=6, kernel_size=3, stride=1, padding=0)
# 前向传播返回x
def forward(self, x):
x = self.conv1(x)
return x
# 神经网络初始化
mynn = Mynn()
# 查看网络结构
print(mynn)
writer = SummaryWriter("logs")
step = 0
# 将数据集每张图像放入神经网络
for data in dataloader:
imgs, traget = data
output = mynn(imgs)
print(imgs.shape)
print(output.shape)
writer.add_images("输入", imgs, step) # 输入(64,3,32,32)
# 由于6通道不知道咋输出,将输出reshape一下。
output = torch.reshape(output, (-1, 3, 30, 30)) # 此处-1为不知道该多少让它自动计算
writer.add_images("输出", output, step) # 输出(64,6,30,30)
step += 1
可以看到3通道变为6通道。
神经网络-最大池化的使用
最大池化有时候也称为下采样。
步长为kernel_size???
ceil_mode为True表示数据不足卷积核的时候,仍可以取最大,如图所示。
# 最大池化
import torch
from torch import nn
from torch.nn import MaxPool2d
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]], dtype=float) # 注意最大池化需要浮点数
# 改变tensor形状
input = torch.reshape(input, (-1, 1, 5, 5))
print(input.shape)
class Tudui(nn.Module):
def __init__(self):
super(Tudui, self).__init__()
self.maxpool1 = MaxPool2d(kernel_size=3, ceil_mode=True)
def forward(self, input):
output = self.maxpool1(input)
return output
tudui = Tudui()
output = tudui(input)
print(output)
当把ceil_mode设为False时,结果为:
为什么要进行最大池化,最大池化作用是什么?
保留输入特征的同时,使数据量减小。对于网络来讲,保留特征数据量减小对网络计算速度有好处。类似于1080P视频最大池化后变为720P,保留内容的同时文件缩小。
那么用图片来试试,直观感受一下,代码如下:
# 最大池化图片
import torch
import torchvision
from torch import nn
from torch.nn import MaxPool2d
from torch.utils.data import DataLoader
from torch.utils.tensorboard import SummaryWriter
cifar_path = "D:\\Scientific_Research\\Pycharmproject\\Biji\\dataset"
dataset = torchvision.datasets.CIFAR10(root=cifar_path, train=False, transform=torchvision.transforms.ToTensor(),
download=True)
dataloader = DataLoader(dataset=dataset, batch_size=64)
class Tudui(nn.Module):
def __init__(self):
super(Tudui, self).__init__()
self.maxpool1 = MaxPool2d(kernel_size=3, ceil_mode=False)
def forward(self, input):
output = self.maxpool1(input)
return output
tudui = Tudui()
writer = SummaryWriter("logs")
step = 0
# 在dataloarder中取数据
for data in dataloader:
imgs, targets = data
writer.add_images("原始", imgs, step)
output = tudui(imgs)
writer.add_images("最大池化", output, step)
step = step + 1
效果就是这样:
神经网络-非线性激活
为神经网络中引入一些非线性特质。
以RELU代码为例:
# RELU
import torch
from torch import nn
from torch.nn import ReLU
input = torch.tensor([[1, -0.5],
[-1, 3]])
input = torch.reshape(input, (-1, 1, 2, 2))
print(input.shape)
class Tudui(nn.Module):
def __init__(self):
super(Tudui, self).__init__()
self.relu1 = ReLU() # 此处inplace设置是否对原来结果进行替换,常选False(默认),防止原始数据丢失
def forward(self,input):
output=self.relu1(input)
return output
tudui=Tudui()
output = tudui(input)
print(output)
输出:
将-1,-0.5变为了0
RELU对图像不明显,试一下sigmoid,代码如下:
# sigmoid
import torch
import torchvision
from torch import nn
from torch.nn import MaxPool2d, ReLU, Sigmoid
from torch.utils.data import DataLoader
from torch.utils.tensorboard import SummaryWriter
cifar_path = "D:\\Scientific_Research\\Pycharmproject\\Biji\\dataset"
dataset = torchvision.datasets.CIFAR10(root=cifar_path, train=False, transform=torchvision.transforms.ToTensor(),
download=True)
dataloader = DataLoader(dataset=dataset, batch_size=64)
class Tudui(nn.Module):
def __init__(self):
super(Tudui, self).__init__()
self.relu1 = ReLU() # 此处inplace设置是否对原来结果进行替换,常选False(默认),防止原始数据丢失
self.sigmoid1 = Sigmoid()
def forward(self, input):
output = self.sigmoid1(input)
return output
tudui = Tudui()
writer = SummaryWriter("logs")
step = 0
for data in dataloader:
imgs, targets = data
writer.add_images("输入", imgs, step)
output = tudui(imgs)
writer.add_images("输出", output, step)
step = step + 1
经过sigmoid变换的图像:
非线性变化主要目的是为网络引入更多非线性特征。
神经网络-线性层及其他层介绍
import torch
import torchvision
from torch import nn
from torch.nn import Linear
from torch.utils.data import DataLoader
cifar_path = "D:\\Scientific_Research\\Pycharmproject\\Biji\\dataset"
dataset = torchvision.datasets.CIFAR10(root=cifar_path, 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.linear1 = Linear(196608, 10)
def forward(self, input):
output = self.linear1(input)
return output
tudui = Tudui()
for data in dataloader:
imgs, targets = data
print(imgs.shape)
# output = torch.reshape(imgs, (1, 1, 1, -1))
output = torch.flatten(imgs) # 展平
print(output.shape)
output = tudui(output)
print(output.shape)
在官方文档中,TORCHVISION.MODELS中提供了一些写好的Module。