如何使用 libtorch 实现 AlexNet 网络?

按照图片上流程写即可。输入的图片大小必须 227x227 3 通道彩色图片

// Define a new Module.
struct Net : torch::nn::Module {
Net() {
conv1 = torch::nn::Conv2d(torch::nn::Conv2dOptions(3, 96, { 11,11 }).stride({4,4}));
conv2 = torch::nn::Conv2d(torch::nn::Conv2dOptions(96, 256, { 5,5 }).padding(2));
conv3 = torch::nn::Conv2d(torch::nn::Conv2dOptions(256, 384, { 3,3 }).padding(1));
conv4 = torch::nn::Conv2d(torch::nn::Conv2dOptions(384, 384, { 3,3 }).padding(1));
conv5 = torch::nn::Conv2d(torch::nn::Conv2dOptions(384, 256, { 3,3 }).padding(1)); fc1 = torch::nn::Linear(256*6*6,4096);
fc2 = torch::nn::Linear(4096, 4096);
fc3 = torch::nn::Linear(4096, 1000);
} // Implement the Net's algorithm.
torch::Tensor forward(torch::Tensor x) { x = conv1->forward(x);
x = torch::relu(x);
//LRN
x = torch::max_pool2d(x, { 3,3 }, { 2,2 });
x = conv2->forward(x);
//LRN
x = torch::relu(x);
x = torch::max_pool2d(x, { 3,3 }, { 2,2 });
x = conv3->forward(x);
x = torch::relu(x);
x = conv4->forward(x);
x = torch::relu(x);
x = conv5->forward(x);
x = torch::relu(x);
x = torch::max_pool2d(x, { 3,3 }, { 2,2 }); x = x.view({ x.size(0),-1 });
x = fc1->forward(x);
x = torch::relu(x);
x = torch::dropout(x,0.5,is_training()); x = fc2->forward(x);
x = torch::relu(x);
x = torch::dropout(x, 0.5, is_training()); x = fc3->forward(x); x = torch::log_softmax(x,1);
return x;
} // Use one of many "standard library" modules.
torch::nn::Conv2d conv1{ nullptr };
torch::nn::Conv2d conv2{ nullptr };
torch::nn::Conv2d conv3{ nullptr };
torch::nn::Conv2d conv4{ nullptr };
torch::nn::Conv2d conv5{ nullptr };
torch::nn::Linear fc1{ nullptr };
torch::nn::Linear fc2{ nullptr };
torch::nn::Linear fc3{ nullptr };
};

具体可参考这个

name: "AlexNet"
layer {
name: "data"
type: "Input"
top: "data"
input_param { shape: { dim: 10 dim: 3 dim: 227 dim: 227 } }
}
layer {
name: "conv1"
type: "Convolution"
bottom: "data"
top: "conv1"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 96
kernel_size: 11
stride: 4
}
}
layer {
name: "relu1"
type: "ReLU"
bottom: "conv1"
top: "conv1"
}
layer {
name: "norm1"
type: "LRN"
bottom: "conv1"
top: "norm1"
lrn_param {
local_size: 5
alpha: 0.0001
beta: 0.75
}
}
layer {
name: "pool1"
type: "Pooling"
bottom: "norm1"
top: "pool1"
pooling_param {
pool: MAX
kernel_size: 3
stride: 2
}
}
layer {
name: "conv2"
type: "Convolution"
bottom: "pool1"
top: "conv2"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 2
kernel_size: 5
group: 2
}
}
layer {
name: "relu2"
type: "ReLU"
bottom: "conv2"
top: "conv2"
}
layer {
name: "norm2"
type: "LRN"
bottom: "conv2"
top: "norm2"
lrn_param {
local_size: 5
alpha: 0.0001
beta: 0.75
}
}
layer {
name: "pool2"
type: "Pooling"
bottom: "norm2"
top: "pool2"
pooling_param {
pool: MAX
kernel_size: 3
stride: 2
}
}
layer {
name: "conv3"
type: "Convolution"
bottom: "pool2"
top: "conv3"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 384
pad: 1
kernel_size: 3
}
}
layer {
name: "relu3"
type: "ReLU"
bottom: "conv3"
top: "conv3"
}
layer {
name: "conv4"
type: "Convolution"
bottom: "conv3"
top: "conv4"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 384
pad: 1
kernel_size: 3
group: 2
}
}
layer {
name: "relu4"
type: "ReLU"
bottom: "conv4"
top: "conv4"
}
layer {
name: "conv5"
type: "Convolution"
bottom: "conv4"
top: "conv5"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 1
kernel_size: 3
group: 2
}
}
layer {
name: "relu5"
type: "ReLU"
bottom: "conv5"
top: "conv5"
}
layer {
name: "pool5"
type: "Pooling"
bottom: "conv5"
top: "pool5"
pooling_param {
pool: MAX
kernel_size: 3
stride: 2
}
}
layer {
name: "fc6"
type: "InnerProduct"
bottom: "pool5"
top: "fc6"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
inner_product_param {
num_output: 4096
}
}
layer {
name: "relu6"
type: "ReLU"
bottom: "fc6"
top: "fc6"
}
layer {
name: "drop6"
type: "Dropout"
bottom: "fc6"
top: "fc6"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "fc7"
type: "InnerProduct"
bottom: "fc6"
top: "fc7"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
inner_product_param {
num_output: 4096
}
}
layer {
name: "relu7"
type: "ReLU"
bottom: "fc7"
top: "fc7"
}
layer {
name: "drop7"
type: "Dropout"
bottom: "fc7"
top: "fc7"
dropout_param {
dropout_ratio: 0.5
}
}
layer {
name: "fc8"
type: "InnerProduct"
bottom: "fc7"
top: "fc8"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
inner_product_param {
num_output: 1000
}
}
layer {
name: "prob"
type: "Softmax"
bottom: "fc8"
top: "prob"
}

如何使用 libtorch 实现 AlexNet 网络?的更多相关文章

  1. AlexNet 网络详解及Tensorflow实现源码

    版权声明:本文为博主原创文章,未经博主允许不得转载. 1. 图片数据处理 2. 卷积神经网络 2.1. 卷积层 2.2. 池化层 2.3. 全链层 3. AlexNet 4. 用Tensorflow搭 ...

  2. 第十六节,卷积神经网络之AlexNet网络实现(六)

    上一节内容已经详细介绍了AlexNet的网络结构.这节主要通过Tensorflow来实现AlexNet. 这里做测试我们使用的是CIFAR-10数据集介绍数据集,关于该数据集的具体信息可以通过以下链接 ...

  3. 第十五节,卷积神经网络之AlexNet网络详解(五)

    原文 ImageNet Classification with Deep ConvolutionalNeural Networks 下载地址:http://papers.nips.cc/paper/4 ...

  4. Caffe训练AlexNet网络,精度不高或者为0的问题结果

    当我们使用Caffe训练AlexNet网络时,会遇到精度一值在低精度(30%左右)升不上去,或者精度总是为0,如下图所示: 出现这种情况,可以尝试使用以下几个方法解决: 1.数据样本量是否太少,最起码 ...

  5. 如何使用 libtorch 实现 LeNet 网络?

    如何使用 libtorch 实现 LeNet 网络? LeNet 网络论文地址: http://yann.lecun.com/exdb/publis/pdf/lecun-01a.pdf

  6. AlexNet网络

    AlexNet 中包含了比较新的技术点,首次在CNN中成功应用了 ReLu .Dropout和LRN等Trick. 1.成功使用了Relu作为CNN的激活函数,并验证其效果在较深的网络中超过了Sigm ...

  7. AlexNet网络的Pytorch实现

    1.文章原文地址 ImageNet Classification with Deep Convolutional Neural Networks 2.文章摘要 我们训练了一个大型的深度卷积神经网络用于 ...

  8. 深入理解AlexNet网络

    原文地址:https://blog.csdn.net/luoluonuoyasuolong/article/details/81750190 AlexNet论文:<ImageNet Classi ...

  9. pytorch实现AlexNet网络

    直接上图吧 写网络就像搭积木

随机推荐

  1. mysql5.5和5.6版本间的坑

    mysql 5.5 int类型 设置不为null,无填充,添加新数据会自动填充0 而5.6同样的配置新建数据没值时,不让添加 5.5 datetime 不能设置默认时间(可以通过某些复杂的方式,这里说 ...

  2. C - The C Answer (2nd Edition) - Exercise 1-2

    /* Experiment to find out what happens when printf's argument string contains \c, where c is some ch ...

  3. am335x 无屏实现开关机程序

    因测试需要加入开机次数记录,所以记录一下7816开关机是怎么做的 原理很简单,开机时判断一个记录文件是否存在,如果存在,运行一段代码,将记录开机次数文件的值读出来+1 代码如下: #include & ...

  4. php7 安装扩展

    安装mongodb扩展 前提:安装好php7,位置:/usr/local/php 方法一: /usr/local/php7/bin/pecl install mongodb service php-f ...

  5. 欧拉函数 & 【POJ】2478 Farey Sequence & 【HDU】2824 The Euler function

    http://poj.org/problem?id=2478 http://acm.hdu.edu.cn/showproblem.php?pid=2824 欧拉函数模板裸题,有两种方法求出所有的欧拉函 ...

  6. ERROR getting 'android:label' attribute: attribute is not a string value

    这个的原因很多地方都有描述,原因多半是多国语言string.xml 有的有这个值, 有的没有. 关键是怎么定位? 实际上他报错的是当前处理的xml element有问题, 而出错的时候盖住了要处理的. ...

  7. 漫游Kafka实战篇之搭建Kafka运行环境(2)

    接下来一步一步搭建Kafka运行环境. Step 1: 下载Kafka 点击下载最新的版本并解压. > tar -xzf kafka_2.9.2-0.8.1.1.tgz > cd kafk ...

  8. iOS TabBar添加阴影

    效果图如下所示: 直接上代码 //移除顶部线条 self.tabBar.backgroundImage = [UIImage new]; self.tabBar.shadowImage = [UIIm ...

  9. Angular ViewChild

    viewchild // 使用方法 git clone https://git.oschina.net/mumu-osc/learn-component.git cd learn-component ...

  10. WPF Expander控件(扩展面板)

    这算是我比较喜欢的一个控件,以前在Winform中也常用类似的.它包装了一块内容,通过单击一个小箭头按钮可以显示或隐藏所包含的内容.在线帮助以及Web页面经常使用这种技术,因为既可以包含大量内容,而又 ...