PyTorch实现AlexNet示例

时间:2022-09-05 11:15:57

PyTorch: https://github.com/shanglianlm0525/PyTorch-Networks

PyTorch实现AlexNet示例

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
import torch
import torch.nn as nn
import torchvision
 
class AlexNet(nn.Module):
  def __init__(self,num_classes=1000):
    super(AlexNet,self).__init__()
    self.feature_extraction = nn.Sequential(
      nn.Conv2d(in_channels=3,out_channels=96,kernel_size=11,stride=4,padding=2,bias=False),
      nn.ReLU(inplace=True),
      nn.MaxPool2d(kernel_size=3,stride=2,padding=0),
      nn.Conv2d(in_channels=96,out_channels=192,kernel_size=5,stride=1,padding=2,bias=False),
      nn.ReLU(inplace=True),
      nn.MaxPool2d(kernel_size=3,stride=2,padding=0),
      nn.Conv2d(in_channels=192,out_channels=384,kernel_size=3,stride=1,padding=1,bias=False),
      nn.ReLU(inplace=True),
      nn.Conv2d(in_channels=384,out_channels=256,kernel_size=3,stride=1,padding=1,bias=False),
      nn.ReLU(inplace=True),
      nn.Conv2d(in_channels=256,out_channels=256,kernel_size=3,stride=1,padding=1,bias=False),
      nn.ReLU(inplace=True),
      nn.MaxPool2d(kernel_size=3, stride=2, padding=0),
    )
    self.classifier = nn.Sequential(
      nn.Dropout(p=0.5),
      nn.Linear(in_features=256*6*6,out_features=4096),
      nn.ReLU(inplace=True),
      nn.Dropout(p=0.5),
      nn.Linear(in_features=4096, out_features=4096),
      nn.ReLU(inplace=True),
      nn.Linear(in_features=4096, out_features=num_classes),
    )
  def forward(self,x):
    x = self.feature_extraction(x)
    x = x.view(x.size(0),256*6*6)
    x = self.classifier(x)
    return x
 
 
if __name__ =='__main__':
  # model = torchvision.models.AlexNet()
  model = AlexNet()
  print(model)
 
  input = torch.randn(8,3,224,224)
  out = model(input)
  print(out.shape)

以上这篇PyTorch实现AlexNet示例就是小编分享给大家的全部内容了,希望能给大家一个参考,也希望大家多多支持服务器之家。

原文链接:https://blog.csdn.net/shanglianlm/article/details/86424857