【深度学习】tensorboard的使用

目前正在写一个训练框架，需要有以下几个功能：
1.保存模型
2.断点继续训练
3.加载模型
4.tensorboard 查询训练记录的功能
命令：
tensorboard --logdir=runs --host=192.168.112.5
效果：

import torch
import torch.nn as nn
import torch.optim as optim
from torch.utils.tensorboard import SummaryWriter# 假设有一个训练数据集和其对应的标签
# 例如：
data = torch.tensor([[1.0], [2.0], [3.0]])
labels = torch.tensor([[2.0], [4.0], [6.0]])class MYdataset(torch.utils.data.Dataset):def __init__(self, features, labels):self.features = featuresself.labels = labelsdef __getitem__(self, index):return self.features[index], self.labels[index]def __len__(self):return len(self.features)BATCH_SIZE= 1train_ds = MYdataset(data, labels)train_dl = torch.utils.data.DataLoader(train_ds,batch_size=BATCH_SIZE,shuffle = True
)# 定义模型、损失函数和优化器
model = nn.Linear(1, 1)
criterion = nn.MSELoss()
optimizer = optim.SGD(model.parameters(), lr=0.01)# 创建summary writer
writer = SummaryWriter('runs/loss_visualization')for epoch in range(10):  # 假设我们只训练10个周期for i, (data,labels) in enumerate(train_dl):  # 假设train_loader是数据加载器# 前向传播outputs = model(data)loss = criterion(outputs, labels)# 反向传播optimizer.zero_grad()loss.backward()optimizer.step()# 将损失记录到TensorBoard中print("Loss:",loss.item())writer.add_scalar('Loss', loss.item(), epoch * len(train_dl) + i)# 关闭summary writer
writer.close()