构建经典PyTorch框架卷积神经网络参数demo
构建经典PyTorch框架卷积神经网络参数demo
构建经典PyTorch框架卷积神经网络参数demo
- 构建经典PyTorch框架卷积神经网络参数demo
- 前言
- 一、定义超参数
- 二、数据集
- 三、定义卷积网络模型
- 四、 验证准确率
- 五、 完整代码实现
- 总结
前言
一、定义超参数
-
图像的总尺寸 28 * 28
input_size = 28; -
标签的种类数
num_classses = 10; -
训练的总循环周期
num_epochs = 3; -
一个最（批次）的大小 64张图片
batch_size = 64;
二、数据集
# 训练集
train_dataset = datasets.MNIST('data/', download=True, train=True, transform=transforms.ToTensor());# 测试集
test_dataset = datasets.MNIST('data/', download=True, train=False, transform=transforms.ToTensor());# 构建batch数据train_loader = torch.utils.data.DataLoader(dataset = train_dataset,batch_size = batch_size,shuffle=True);test_loader = torch.utils.data.DataLoader(dataset = test_dataset,batch_size = batch_size,shuffle=True);
三、定义卷积网络模型
- Conv2d
- ReLU
- 池化
class CNN(nn.Module):def __init__(self):super(CNN, self).__init__();self.conv1 = nn.Sequential( # 输入大小 （1， 28， 28）# 2D卷积9nn.Conv2d(in_channels = 1, # 灰度图out_channels = 16, # 要得到几多少个特征图kernel_size = 5, # 卷积核大小stride = 1, # 步长padding = 2, # 如果希望卷积后大小更原来一样 需要设置padding=(kernel_size-1)/2 if stride = 1), # 输出的特征图为（16， 28， 28）nn.ReLU(), # relu层nn.MaxPool2d(kernel_size = 2), # 进行池化操作（2*2区域） 输出结果为（16， 14， 14););self.conv2 = nn.Sequential( # 下一个套餐的输入(16, 14, 14)nn.Conv2d(16, 32, 5, 1, 2), # 输出（32， 14， 14）nn.ReLU(), # relu层nn.Conv2d(32, 32, 5, 1, 2), # nn.ReLU(),nn.MaxPool2d(2), # 输出（32， 7， 7）);self.conv3 = nn.Sequential( # 下一个套餐的输入(16, 14, 14)nn.Conv2d(32, 64, 5, 1, 2), # 输出（32， 14， 14）nn.ReLU(), # 输出 （32， 7， 7）);self.out = nn.Linear(64 * 7 * 7, 10); # 全连接层得到的结果# 前向传播def forward(self, x):x = self.conv1(x);x = self.conv2(x);x = self.conv3(x);x = x.view(x.size(0), -1); # flatten 操作 结果为：（batch_size， 32 * 7 * 7）output = self.out(x); return output;四、 验证准确率
# 实例化net = CNN();# 损失函数
criterion = nn.CrossEntropyLoss();# 优化器 定义优化器 普通的随机梯度下降算法 lr : 学习率
optimizer = optim.Adam(net.parameters(), lr=0.001);# 开始训练循环
for epoch in range(num_epochs):# 当前epoch的结果保存下来train_rights = [];#针对容器中的每一个批进行循环for batch_idx, (data, target) in enumerate(train_loader): net.train();output = net(data);loss = criterion(output, target);optimizer.zero_grad();loss.backward();optimizer.step();right = accuracy(output, target);train_rights.append(right);if batch_idx % 100 == 0:net.eval();val_rights = [];for (data, target) in test_loader:output = net(data);right = accuracy(output, target);val_rights.append(right);# 准确率计算train_r = (sum([tup[0] for tup in train_rights]), sum([tup[1] for tup in train_rights]));val_r = (sum([tup[0] for tup in val_rights]), sum([tup[1] for tup in val_rights]));print('当前epoch：{} [{}/{} ({:.0f}%)]\t 损失: {:.6f}\t 训练集准确率:{:.2f}%\t 测试集准确率:{:.2f}%'.format(epoch, batch_idx * batch_size, len(train_loader.dataset),100. * batch_idx / len(train_loader),loss.data,100. *train_r[0].numpy() / train_r[1],100. * val_r[0].numpy() / val_r[1]));
五、 完整代码实现
## 基本环境import torch;
import torch.nn as nn;
import torch.optim as optim;
import torch.nn.functional as F;
# datasets 处理数据集
# transforms 预处理操作的
from torchvision import datasets, transforms;
import matplotlib.pyplot as plt;
import numpy as np;#%matplotlib inline# 定义超参数# 1. 图像的总尺寸 28 * 28
input_size = 28;
# 2. 标签的种类数
num_classses = 10;# 3. 训练的总循环周期
num_epochs = 3;
# 4. 一个最（批次）的大小 64张图片
batch_size = 64;# 训练集
train_dataset = datasets.MNIST('data/', download=True, train=True, transform=transforms.ToTensor());# 测试集
test_dataset = datasets.MNIST('data/', download=True, train=False, transform=transforms.ToTensor());# 构建batch数据train_loader = torch.utils.data.DataLoader(dataset = train_dataset,batch_size = batch_size,shuffle=True);test_loader = torch.utils.data.DataLoader(dataset = test_dataset,batch_size = batch_size,shuffle=True);# 定义卷积网络模型class CNN(nn.Module):def __init__(self):super(CNN, self).__init__();self.conv1 = nn.Sequential( # 输入大小 （1， 28， 28）# 2D卷积9nn.Conv2d(in_channels = 1, # 灰度图out_channels = 16, # 要得到几多少个特征图kernel_size = 5, # 卷积核大小stride = 1, # 步长padding = 2, # 如果希望卷积后大小更原来一样 需要设置padding=(kernel_size-1)/2 if stride = 1), # 输出的特征图为（16， 28， 28）nn.ReLU(), # relu层nn.MaxPool2d(kernel_size = 2), # 进行池化操作（2*2区域） 输出结果为（16， 14， 14););self.conv2 = nn.Sequential( # 下一个套餐的输入(16, 14, 14)nn.Conv2d(16, 32, 5, 1, 2), # 输出（32， 14， 14）nn.ReLU(), # relu层nn.Conv2d(32, 32, 5, 1, 2), # nn.ReLU(),nn.MaxPool2d(2), # 输出（32， 7， 7）);self.conv3 = nn.Sequential( # 下一个套餐的输入(16, 14, 14)nn.Conv2d(32, 64, 5, 1, 2), # 输出（32， 14， 14）nn.ReLU(), # 输出 （32， 7， 7）);self.out = nn.Linear(64 * 7 * 7, 10); # 全连接层得到的结果# 前向传播def forward(self, x):x = self.conv1(x);x = self.conv2(x);x = self.conv3(x);x = x.view(x.size(0), -1); # flatten 操作 结果为：（batch_size， 32 * 7 * 7）output = self.out(x); return output;# 准确率def accuracy(predictions, labels):pred = torch.max(predictions.data, 1)[1];# 预测值与标签是否相等的rights = pred.eq(labels.data.view_as(pred)).sum();return rights, len(labels);# 实例化net = CNN();# 损失函数
criterion = nn.CrossEntropyLoss();# 优化器 定义优化器 普通的随机梯度下降算法 lr : 学习率
optimizer = optim.Adam(net.parameters(), lr=0.001);# 开始训练循环
for epoch in range(num_epochs):# 当前epoch的结果保存下来train_rights = [];#针对容器中的每一个批进行循环for batch_idx, (data, target) in enumerate(train_loader): net.train();output = net(data);loss = criterion(output, target);optimizer.zero_grad();loss.backward();optimizer.step();right = accuracy(output, target);train_rights.append(right);if batch_idx % 100 == 0:net.eval();val_rights = [];for (data, target) in test_loader:output = net(data);right = accuracy(output, target);val_rights.append(right);# 准确率计算train_r = (sum([tup[0] for tup in train_rights]), sum([tup[1] for tup in train_rights]));val_r = (sum([tup[0] for tup in val_rights]), sum([tup[1] for tup in val_rights]));print('当前epoch：{} [{}/{} ({:.0f}%)]\t 损失: {:.6f}\t 训练集准确率:{:.2f}%\t 测试集准确率:{:.2f}%'.format(epoch, batch_idx * batch_size, len(train_loader.dataset),100. * batch_idx / len(train_loader),loss.data,100. *train_r[0].numpy() / train_r[1],100. * val_r[0].numpy() / val_r[1]));运行效果图
总结
人工智能深度学习：https://github.com/chensongpoixs/cdeeplearning_pytorch