泰州网站建设方案推广,海网站建设,网站建设的总体目标考核指标,腾讯邮箱企业邮箱注册目录 1、新建模型 train_model.py
2、运行模型
#xff08;1#xff09;首先会下载data文件库
#xff08;2#xff09;完成之后会开始训练模型#xff08;10次#xff09;
3、 训练好之后#xff0c;进入命令集 4、输入命令#xff1a;python -m tensorboard.ma…目录 1、新建模型 train_model.py
2、运行模型
1首先会下载data文件库
2完成之后会开始训练模型10次
3、 训练好之后进入命令集 4、输入命令python -m tensorboard.main --logdirC:\Users\15535\Desktop\day6\train
1目录的绝对路径获得方法 5、打开网页可视化图形
1运行完之后会自动有一个网址点进去 2显示 1、新建模型 train_model.py
import torch
import torchvision.transforms
from torch.utils.tensorboard import SummaryWriter
from torchvision import datasets
from torch.utils.data import DataLoader
import torch.nn as nn
from torch.nn import CrossEntropyLoss#step1.下载数据集train_datadatasets.CIFAR10(./data,trainTrue,\transformtorchvision.transforms.ToTensor(),downloadTrue)
test_datadatasets.CIFAR10(./data,trainFalse,\transformtorchvision.transforms.ToTensor(),downloadTrue)print(len(train_data))
print(len(test_data))#step2.数据集打包
train_data_loaderDataLoader(train_data,batch_size64,shuffleFalse)
test_data_loaderDataLoader(test_data,batch_size64,shuffleFalse)#step3.搭建网络模型class My_Module(nn.Module):def __init__(self):super(My_Module,self).__init__()#64*32*32*32self.conv1nn.Conv2d(in_channels3,out_channels32,\kernel_size5,padding2)#64*32*16*16self.maxpool1nn.MaxPool2d(2)#64*32*16*16self.conv2nn.Conv2d(in_channels32,out_channels32,\kernel_size5,padding2)#64*32*8*8self.maxpool2nn.MaxPool2d(2)#64*64*8*8self.conv3nn.Conv2d(in_channels32,out_channels64,\kernel_size5,padding2)#64*64*4*4self.maxpool3nn.MaxPool2d(2)#线性化self.flattennn.Flatten()self.linear1nn.Linear(in_features1024,out_features64)self.linear2nn.Linear(in_features64,out_features10)def forward(self,input):#input:64,3,32,32output1self.conv1(input)output2self.maxpool1(output1)output3self.conv2(output2)output4self.maxpool2(output3)output5self.conv3(output4)output6self.maxpool3(output5)output7self.flatten(output6)output8self.linear1(output7)output9self.linear2(output8)return output9my_modelMy_Module()
# print(my_model)
loss_funcCrossEntropyLoss()#衡量模型训练的过程输入输出之间的差值
#优化器lr越大模型就越“聪明”
optim torch.optim.SGD(my_model.parameters(),lr0.001)writerSummaryWriter(./train)
#################################训练###############################
for looptime in range(10): #模型训练的次数10print(------looptime:{}------.format(looptime1))num0loss_all0for data in (train_data_loader):num1#前向imgs, targets dataoutput my_model(imgs)loss_train loss_func(output,targets)loss_allloss_allloss_trainif num%1000:print(loss_train)#后向backward 三步法 获取最小的损失函数optim.zero_grad()loss_train.backward()optim.step()# print(output.shape)loss_avloss_all/len(test_data_loader)print(loss_av)writer.add_scalar(train_loss,loss_av,looptime)writer.close()
#################################验证#########################with torch.no_grad():accuracy0test_loss_all0for data in test_data_loader:imgs,targets dataoutput my_model(imgs)loss_test loss_func(output,targets)#output.argmax(1)---输出标签accuracy(output.argmax(1)targets).sum()test_loss_all test_loss_allloss_testtest_loss_av test_loss_all/len(test_data_loader)acc_av accuracy/len(test_data_loader)print(测试集的平均损失{}测试集的准确率{}.format(test_loss_av,acc_av))writer.add_scalar(test_loss,test_loss_av,looptime)writer.add_scalar(acc,acc_av,looptime)writer.close()
2、运行模型
1首先会下载data文件库
2完成之后会开始训练模型10次 3、 训练好之后进入命令集 4、输入命令python -m tensorboard.main --logdirC:\Users\15535\Desktop\day6\train 1目录的绝对路径获得方法
执行下面的操作自动复制 5、打开网页可视化图形
1运行完之后会自动有一个网址点进去 2显示
