Tensorboard可视化

tensorboard 可视化神经网络 

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


class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(in_channels=1,out_channels=10,kernel_size=5)
        self.conv2 = nn.Conv2d(in_channels=10,out_channels=20,kernel_size=5)
        self.conv2_drop = nn.Dropout2d()    # 维度不变，只是把通道C里的数据按概率赋值为0
        self.fc1 = nn.Linear(in_features=320,out_features=50)
        self.fc2 = nn.Linear(in_features=50,out_features=10)
        self.bn = nn.BatchNorm2d(num_features=20)    # C = 20
        # Batch Normalization强行将数据拉回到均值为0，方差为1的正太分布上，一方面使得数据分布一致，另一方面避免梯度消失。

    def forward(self,x):
        x = F.max_pool2d(self.conv1(x),2)
        x = F.relu(x) + F.relu(-x)
        x = F.relu(F.max_pool2d(self.conv2_drop(self.conv2(x)),kernel_size=2))
        x = self.bn(x)
        x = x.view(-1,320)
        x = F.relu(self.fc1(x))
        x = F.dropout(x,training=self.training)   # 随机采样，防止过拟合
        x = self.fc2(x)
        x = F.softmax(x,dim=1)
        return x
input = torch.rand(32,1,28,28)                    # (N样本数量,C图像通道数,W宽,H高)
model = Net()
writer = SummaryWriter('tensorb')
writer.add_graph(model,(input,))
writer.close()

在Terminal终端打开输入 tensorboard --logdir=tensorb

 tensorboard 可视化损失值

import numpy as np
import torch
import torch.nn as nn
from torch.utils.tensorboard import SummaryWriter

dtype = torch.FloatTensor
writer = SummaryWriter('loss')
np.random.seed(100)
x_train = np.linspace(-1,1,100).reshape(100,1)
y_train = 3 * np.power(x_train,2) + 2 + 0.2*np.random.rand(x_train.size).reshape(100,1)
model = nn.Linear(1,100)
criterion = nn.MSELoss()
optimizer = torch.optim.SGD(model.parameters(),lr = 1)
for epoch in range(100):
    inputs = torch.from_numpy(x_train).type(dtype)
    targets = torch.from_numpy(y_train).type(dtype)
    output = model(inputs)
    loss = criterion(output,targets)
    optimizer.zero_grad()
    loss.backward()
    optimizer.step()
    writer.add_scalar('训练损失值',loss,epoch)