使用tensorboard查看模型

1、安装

在pytorch的环境下，安装tensorboard和tensorflow

conda install tensorboard
conda install tensorflow

我安装在虚拟环境d2l中 

验证是否安装成功

from torch.utils.tensorboard import SummaryWriter

输入如上代码，如果没有报错，则代表安装成功

2、简单实例

在pytorch官网找到实例

创建日志 

from torch.utils.tensorboard import SummaryWriter
#创建1个名为“logs”的日志
writer = SummaryWriter("logs")
x = range(100)
for i in x:
    #写入y=2*x的曲线
    writer.add_scalar('y=2x', i * 2, i)
#关闭
writer.close()

运行结束后，已经创建了1个日志文件 

消费日志

打开终端控制台 

 在控制台，检查是否路径正确。正确的话，输入以下，并回车。logs是日志名

 tensorboard --logdir=logs

点击链接，查看即可 

注意：不能同时查看多个日志，容易出错

3、画模型函数--以resnet为例

from torch.utils.tensorboard import SummaryWriter
import d2l.torch
import torch
from torch.nn import functional as F
from torch import nn

class Residual(nn.Module): #@save
    def __init__(self,input_channels,num_channels,use_1x1conv=False,strides=1):
        super().__init__()      self.conv1=nn.Conv2d(input_channels,num_channels,kernel_size=3,padding=1,stride=strides)
        self.conv2=nn.Conv2d(num_channels,num_channels,kernel_size=3,padding=1)
        if use_1x1conv:
        self.conv3=nn.Conv2d(input_channels,num_channels,kernel_size=1,stride=strides)
        else:
            self.conv3=None
        self.bn1=nn.BatchNorm2d(num_channels)
        self.bn2=nn.BatchNorm2d(num_channels)
    def forward(self,x):
        y=F.relu(self.bn1(self.conv1(x)))
        y=self.bn2(self.conv2(y))
        if self.conv3:
            x=self.conv3(x)
        y+=x
        return F.relu(y)
b1 = nn.Sequential(nn.Conv2d(in_channels=1,out_channels=64,kernel_size=7,padding=3,stride=2),
                   nn.BatchNorm2d(64),
                   nn.ReLU(),
                   nn.MaxPool2d(kernel_size=3,padding=1,stride=2))#conv2d尺寸减半，池化也
def resnet_block(input_channels,num_channels,num_residuls,first_block=False):
    block = []
    for i in range(num_residuls):
        if i==0 and not first_block:
            block.append(
                Residual(input_channels,num_channels,use_1x1conv=True,strides=2))
        else:
            block.append(
                Residual(num_channels,num_channels))
    return block
b2 = nn.Sequential(*resnet_block(64,64,2,True))
b3 = nn.Sequential(*resnet_block(64,128,2,False))
b4 = nn.Sequential(*resnet_block(128,256,2,False))
b5 = nn.Sequential(*resnet_block(256,512,2,False))
resnet = nn.Sequential(b1,b2,b3,b4,b5,
                       nn.AdaptiveAvgPool2d((1,1)),
                       nn.Flatten(),
                       nn.Linear(in_features=512,out_features=10))
#以上是resnet网络的定义

writer = SummaryWriter("model_logs")
#创建1个名为“model_logs”日志
X = torch.randn(size=(1,1,224,224))
writer.add_graph(model=resnet,input_to_model=X)
#将输入X代入模型，写进日志
writer.close()
#关闭日志

成功创建名为“model_logs”的日志 

 

 打开终端，点击本地链接查看网络

会显示下图所示的大致框架 

双击加号之后，会展现出具体的网络细节 ，并且显示了每阶段向量形状。每一步骤都可以继续深化展现