Python Tensorflow Tensorboard 默认端口

声明:本页面是StackOverFlow热门问题的中英对照翻译,遵循CC BY-SA 4.0协议,如果您需要使用它,必须同样遵循CC BY-SA许可,注明原文地址和作者信息,同时你必须将它归于原作者(不是我):StackOverFlow 原文地址: http://stackoverflow.com/questions/35551326/
Warning: these are provided under cc-by-sa 4.0 license. You are free to use/share it, But you must attribute it to the original authors (not me): StackOverFlow

提示:将鼠标放在中文语句上可以显示对应的英文。显示中英文
时间:2020-08-19 16:35:20  来源:igfitidea点击:

Tensorflow Tensorboard default port

pythontensorflowtensorboard

提问by nicolasdavid

Is there a way to change the default port ("6006") on tensorboard so we could open multiple tensorboard? Maybe an option like --port="8008" ?

有没有办法更改张量板上的默认端口(“6006”),以便我们可以打开多个张量板?也许像 --port="8008" 这样的选项?

采纳答案by nicolasdavid

In fact there is an option to change the default port ...

事实上,有一个选项可以更改默认端口...

tensorboard --logdir=/tmp  --port=8008

回答by Salvador Dali

You should provide a port flag (--port=6007).

您应该提供端口标志 ( --port=6007)。

But I am here to explain how you can find it and other flags without any documentation. Almost all command line tools have a flag -hor --helpwhich shows all possible flags this tool allows.

但我在这里解释如何在没有任何文档的情况下找到它和其他标志。几乎所有的命令行工具都有一个标志-h--help显示该工具允许的所有可能的标志。

By running it you will see information about a port flag and that --logdirallows you to

通过运行它,您将看到有关端口标志的信息,并--logdir允许您

also pass a comma separated list of log directories

还传递一个逗号分隔的日志目录列表

and you can also inspect separate event-files and tags with --event_fileand --tagflags

您还可以使用--event_file--tag标志检查单独的事件文件和标签

回答by Miladiouss

If you are already in the directory where TensorFlow writes its logs, you should specify the port first:

如果您已经在 TensorFlow 写入日志的目录中,您应该首先指定端口:

tensorboard --port=6007 --logdir runs

If you are feeding a directory to logdir, then the order doesn't matter. (I am using TensorBaord 1.8)

如果您将目录提供给logdir,则顺序无关紧要。(我使用的是 TensorBaord 1.8)