网站推广方法是什么,wordpress 支持中文用户名,濮阳网络改造,网站免费建设推荐Pytorch从零开始实战——ResNet-50算法实战
本系列来源于365天深度学习训练营
原作者K同学 文章目录 Pytorch从零开始实战——ResNet-50算法实战环境准备数据集模型选择开始训练可视化模型预测总结 环境准备
本文基于Jupyter notebook#xff0c;使用Python3.8#xff0c…Pytorch从零开始实战——ResNet-50算法实战
本系列来源于365天深度学习训练营
原作者K同学 文章目录 Pytorch从零开始实战——ResNet-50算法实战环境准备数据集模型选择开始训练可视化模型预测总结 环境准备
本文基于Jupyter notebook使用Python3.8Pytorch2.0.1cu118torchvision0.15.2需读者自行配置好环境且有一些深度学习理论基础。本次实验的目的是理解ResNet-50模型。 第一步导入常用包
import torch
import torch.nn as nn
import matplotlib.pyplot as plt
import torchvision
import torchvision.transforms as transforms
import torchvision.datasets as datasets
import torch.nn.functional as F
import random
from time import time
import numpy as np
import pandas as pd
import datetime
import gc
import os
import copy
import warnings
os.environ[KMP_DUPLICATE_LIB_OK]True # 用于避免jupyter环境突然关闭
torch.backends.cudnn.benchmarkTrue # 用于加速GPU运算的代码设置随机数种子
torch.manual_seed(428)
torch.cuda.manual_seed(428)
torch.cuda.manual_seed_all(428)
random.seed(428)
np.random.seed(428)检查设备对象
device torch.device(cuda if torch.cuda.is_available() else cpu)
device, torch.cuda.device_count() # # (device(typecuda), 2)数据集
本次数据集是使用鸟的图片分别有四种类别的鸟根据鸟的类别名称存放在不同的文件夹中。 使用pathlib查看类别
import pathlib
data_dir ./data/bird_photos/
data_dir pathlib.Path(data_dir) # 转成pathlib.Path对象
data_paths list(data_dir.glob(*))
classNames [str(path).split(/)[2] for path in data_paths]
classNames # [Black Throated Bushtiti, Cockatoo, Black Skimmer, Bananaquit]使用transforms对数据集进行统一处理并且根据文件夹名映射对应标签
all_transforms transforms.Compose([transforms.Resize([224, 224]),transforms.ToTensor(),transforms.Normalize(mean[0.485, 0.456, 0.406], std[0.229, 0.224, 0.225]) # 标准化
])total_data datasets.ImageFolder(./data/bird_photos/, transformall_transforms)
total_data.class_to_idx# {Bananaquit: 0,# Black Skimmer: 1,# Black Throated Bushtiti: 2,# Cockatoo: 3}随机查看5张图片
def plotsample(data):fig, axs plt.subplots(1, 5, figsize(10, 10)) #建立子图for i in range(5):num random.randint(0, len(data) - 1) #首先选取随机数随机选取五次#抽取数据中对应的图像对象make_grid函数可将任意格式的图像的通道数升为3而不改变图像原始的数据#而展示图像用的imshow函数最常见的输入格式也是3通道npimg torchvision.utils.make_grid(data[num][0]).numpy()nplabel data[num][1] #提取标签 #将图像由(3, weight, height)转化为(weight, height, 3)并放入imshow函数中读取axs[i].imshow(np.transpose(npimg, (1, 2, 0))) axs[i].set_title(nplabel) #给每个子图加上标签axs[i].axis(off) #消除每个子图的坐标轴plotsample(total_data)根据8比2划分数据集和测试集并且利用DataLoader划分批次和随机打乱
train_size int(0.8 * len(total_data))
test_size len(total_data) - train_size
train_ds, test_ds torch.utils.data.random_split(total_data, [train_size, test_size])batch_size 32
train_dl torch.utils.data.DataLoader(train_ds,batch_sizebatch_size,shuffleTrue,)
test_dl torch.utils.data.DataLoader(test_ds,batch_sizebatch_size,shuffleTrue,)len(train_dl.dataset), len(test_dl.dataset) # (452, 113)模型选择
ResNetResidual Network是一种深度神经网络架构由微软亚洲研究院的研究员Kaiming He等人于2015年提出。ResNet的设计主要解决了深度神经网络训练过程中的梯度消失和梯度爆炸等问题使得训练更深的网络变得更为可行。
ResNet的关键创新点是引入了残差学习Residual Learning的概念。在传统的网络中每一层的输入都是由前一层输出直接得到的而ResNet则通过引入残差块Residual Block改变了这种方式。
残差块包含了一个跳跃连接skip connection将输入直接添加到网络的输出形成了残差学习的结构。当输入为x经过一个残差块后的输出为H(x)则残差块的计算方式可以表示为H(x)F(x)x其中F(x)表示残差块的映射函数。由于存在跳跃连接即直接将输入x加到输出上这种结构使得神经网络能够学习恒等映射即H(x)x从而更容易学习到恒等映射的变化部分有助于减轻梯度爆炸或梯度消失问题。
ResNet的整体结构由多个残差块组成包括一些卷积层、批归一化Batch Normalization和非线性激活函数。在深度网络中ResNet的设计使得可以训练非常深的模型。
本次实验的ResNet-50有两个基本的块分别名为Conv Block和Identity Block借用K同学所绘制的图片。 首先构建ResNet中的恒等块恒等块是ResNet中的基本构建模块用于在网络中引入残差学习这个模块将输入x保存一份为tx进行三个卷积最终进行跳跃连接将x和t相加。
class IdentityBlock(nn.Module):def __init__(self, in_channels, out_channels, kernel_size):super(IdentityBlock, self).__init__()self.conv1 nn.Conv2d(in_channels, out_channels[0], kernel_size1)self.bn1 nn.BatchNorm2d(out_channels[0])self.relu1 nn.ReLU()self.conv2 nn.Conv2d(out_channels[0], out_channels[1], kernel_sizekernel_size, padding1)self.bn2 nn.BatchNorm2d(out_channels[1])self.relu2 nn.ReLU()self.conv3 nn.Conv2d(out_channels[1], out_channels[2], kernel_size1)self.bn3 nn.BatchNorm2d(out_channels[2])self.relu4 nn.ReLU()def forward(self, x):t xx self.conv1(x)x self.bn1(x)x self.relu1(x)x self.conv2(x)x self.bn2(x)x self.relu2(x)x self.conv3(x)x self.bn3(x)x tx self.relu4(x)return x下面构建ResNet中的卷积模块与恒等模块类似不过t也经过一次卷积。
class ConvBlock(nn.Module):def __init__(self, in_channels, out_channels, kernel_size, strides(2, 2)):super(ConvBlock, self).__init__()self.conv1 nn.Conv2d(in_channels, out_channels[0], kernel_size1, stridestrides)self.bn1 nn.BatchNorm2d(out_channels[0])self.relu1 nn.ReLU()self.conv2 nn.Conv2d(out_channels[0], out_channels[1], kernel_sizekernel_size, padding1)self.bn2 nn.BatchNorm2d(out_channels[1])self.relu2 nn.ReLU()self.conv3 nn.Conv2d(out_channels[1], out_channels[2], kernel_size1)self.bn3 nn.BatchNorm2d(out_channels[2])self.conv nn.Conv2d(in_channels, out_channels[2], kernel_size1, stridestrides)self.bn nn.BatchNorm2d(out_channels[2])self.relu4 nn.ReLU()def forward(self, x):t xx self.conv1(x)x self.bn1(x)x self.relu1(x)x self.conv2(x)x self.bn2(x)x self.relu2(x)x self.conv3(x)x self.bn3(x)t self.conv(t)t self.bn(t)x tx self.relu4(x)return x构建ResNet50使用上面的恒等块和卷积块能够构建很深的网络模型。
class ResNet50(nn.Module):def __init__(self, input_shape(3, 224, 224), num_classes1000):super(ResNet50, self).__init__()self.conv1 nn.Conv2d(input_shape[0], 64, kernel_size7, stride2, padding3)self.bn_conv1 nn.BatchNorm2d(64)self.relu nn.ReLU()self.maxpool nn.MaxPool2d(kernel_size3, stride2, padding1)self.conv2 ConvBlock(64, [64, 64, 256], kernel_size3, strides(1, 1))self.identity_block1 IdentityBlock(256, [64, 64, 256], kernel_size3)self.identity_block2 IdentityBlock(256, [64, 64, 256], kernel_size3)self.conv3 ConvBlock(256, [128, 128, 512], kernel_size3)self.identity_block3 IdentityBlock(512, [128, 128, 512], kernel_size3)self.identity_block4 IdentityBlock(512, [128, 128, 512], kernel_size3)self.identity_block5 IdentityBlock(512, [128, 128, 512], kernel_size3)self.conv4 ConvBlock(512, [256, 256, 1024], kernel_size3)self.identity_block6 IdentityBlock(1024, [256, 256, 1024], kernel_size3)self.identity_block7 IdentityBlock(1024, [256, 256, 1024], kernel_size3)self.identity_block8 IdentityBlock(1024, [256, 256, 1024], kernel_size3)self.identity_block9 IdentityBlock(1024, [256, 256, 1024], kernel_size3)self.identity_block10 IdentityBlock(1024, [256, 256, 1024], kernel_size3)self.conv5 ConvBlock(1024, [512, 512, 2048], kernel_size3)self.identity_block11 IdentityBlock(2048, [512, 512, 2048], kernel_size3)self.identity_block12 IdentityBlock(2048, [512, 512, 2048], kernel_size3)self.avg_pool nn.AvgPool2d(kernel_size7)self.fc nn.Linear(2048, num_classes)def forward(self, x):x self.conv1(x)x self.bn_conv1(x)x self.relu(x)x self.maxpool(x)x self.conv2(x)x self.identity_block1(x)x self.identity_block2(x)x self.conv3(x)x self.identity_block3(x)x self.identity_block4(x)x self.identity_block5(x)x self.conv4(x)x self.identity_block6(x)x self.identity_block7(x)x self.identity_block8(x)x self.identity_block9(x)x self.identity_block10(x)x self.conv5(x)x self.identity_block11(x)x self.identity_block12(x)x self.avg_pool(x)x x.view(x.size(0), -1)x self.fc(x)return x使用summary查看模型架构
from torchsummary import summary
model ResNet50().to(device)
summary(model, input_size(3, 224, 224))开始训练
定义训练函数
def train(dataloader, model, loss_fn, opt):size len(dataloader.dataset)num_batches len(dataloader)train_acc, train_loss 0, 0for X, y in dataloader:X, y X.to(device), y.to(device)pred model(X)loss loss_fn(pred, y)opt.zero_grad()loss.backward()opt.step()train_acc (pred.argmax(1) y).type(torch.float).sum().item()train_loss loss.item()train_acc / sizetrain_loss / num_batchesreturn train_acc, train_loss定义测试函数
def test(dataloader, model, loss_fn):size len(dataloader.dataset)num_batches len(dataloader)test_acc, test_loss 0, 0with torch.no_grad():for X, y in dataloader:X, y X.to(device), y.to(device)pred model(X)loss loss_fn(pred, y)test_acc (pred.argmax(1) y).type(torch.float).sum().item()test_loss loss.item()test_acc / sizetest_loss / num_batchesreturn test_acc, test_loss定义学习率、损失函数、优化算法
loss_fn nn.CrossEntropyLoss()
learn_rate 0.00001
opt torch.optim.Adam(model.parameters(), lrlearn_rate)开始训练epoch设置为30
import time
epochs 30
train_loss []
train_acc []
test_loss []
test_acc []T1 time.time()best_acc 0
best_model 0for epoch in range(epochs):model.train()epoch_train_acc, epoch_train_loss train(train_dl, model, loss_fn, opt)model.eval() # 确保模型不会进行训练操作epoch_test_acc, epoch_test_loss test(test_dl, model, loss_fn)if epoch_test_acc best_acc:best_acc epoch_test_accbest_model copy.deepcopy(model)train_acc.append(epoch_train_acc)train_loss.append(epoch_train_loss)test_acc.append(epoch_test_acc)test_loss.append(epoch_test_loss)print(epoch:%d, train_acc:%.1f%%, train_loss:%.3f, test_acc:%.1f%%, test_loss:%.3f% (epoch 1, epoch_train_acc * 100, epoch_train_loss, epoch_test_acc * 100, epoch_test_loss))T2 time.time()
print(程序运行时间:%s秒 % (T2 - T1))PATH ./best_model.pth # 保存的参数文件名
if best_model is not None:torch.save(best_model.state_dict(), PATH)print(保存最佳模型)
print(Done)结果稍微有些过拟合
可视化
将训练与测试过程可视化
import warnings
warnings.filterwarnings(ignore) #忽略警告信息
plt.rcParams[font.sans-serif] [SimHei] # 用来正常显示中文标签
plt.rcParams[axes.unicode_minus] False # 用来正常显示负号
plt.rcParams[figure.dpi] 100 #分辨率epochs_range range(epochs)plt.figure(figsize(12, 3))
plt.subplot(1, 2, 1)plt.plot(epochs_range, train_acc, labelTraining Accuracy)
plt.plot(epochs_range, test_acc, labelTest Accuracy)
plt.legend(loclower right)
plt.title(Training and Validation Accuracy)plt.subplot(1, 2, 2)
plt.plot(epochs_range, train_loss, labelTraining Loss)
plt.plot(epochs_range, test_loss, labelTest Loss)
plt.legend(locupper right)
plt.title(Training and Validation Loss)
plt.show()模型预测
定义预测函数
from PIL import Image classes list(total_data.class_to_idx)def predict_one_image(image_path, model, transform, classes):test_img Image.open(image_path).convert(RGB)plt.imshow(test_img) # 展示预测的图片test_img transform(test_img)img test_img.to(device).unsqueeze(0)model.eval()output model(img)_,pred torch.max(output,1)pred_class classes[pred]print(f预测结果是{pred_class})开始预测
predict_one_image(image_path./data/bird_photos/Cockatoo/001.jpg, modelmodel, transformall_transforms, classesclasses) # 预测结果是Cockatoo总结
本次实验学习了ResNet的基本概念和实现ResNet的核心思想是通过引入残差块使网络能够更容易地学习恒等映射的变化部分所以能够构建深层次的网络 同时其中的跳跃连接通过将输入直接添加到输出有助于梯度的流动减轻梯度消失的问题但是ResNet计算和存储的资源要求高容易过拟合也是它的缺点我们可以通过学习它的网络设计思想构建自己的网络。
