PyTorch使用tensorboard可以显示网络运行情况，和TensorFlow的tensorboard使用很类似。均需要安装tensorboard包。
pip install tensorboard==1.15.0导入：
from torch.utils.tensorboard import SummaryWriter指定log路径：
log_dir = './run_logs' self.writer = SummaryWriter(log_dir)写入到log：
self.board_write_step = 50#每50个step写一次board_log if (global_steps+1) % self.board_write_step == 0: self.writer.add_image("result", result_img, global_step=global_steps) self.writer.add_scalar("acc", acc, global_step=global_steps)在shell中开启tensorboard：
tensorboard --logdir=./run_logs浏览器中打开localhost:6006： 注：在新版tensorboard（2.3.0）中，如果需要远程访问，需要在命令中加 --bind_all参数。例如
tensorboard --bind_all --logdir=./run_logs否则只能本机访问，tensorboard启动时会提示这条信息：
Serving TensorBoard on localhost; to expose to the network, use a proxy or pass --bind_all