torch使用tensorboard可视化训练过程

1.导入模块

from torch.utils.tensorboard import SummaryWriter

2.在训练开始之前定义一个writer，括号内为tensorboard保存的文件夹地址：

writer = SummaryWriter('./runs')

3.训练过程中每个epoch写入相应的点：

writer.add_scalar('LOSS/Train_loss', float(avg_train_loss),(epoch + 1))
writer.add_scalar('LOSS/Valid_loss', float(avg_valid_loss), (epoch + 1))
writer.add_scalar('ACC/Train_lacc', float(avg_train_acc), (epoch + 1))
writer.add_scalar('ACC/Valid_lacc', float(avg_valid_acc), (epoch + 1))

4.在终端cd到前面writer中设置的地址：

cd runs/

再输入这个命令：

tensorboard --logdir ./(writer地址)
或：
tensorboard --logdir=./(writer地址)

效果如下：