Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Launching TensorBoard 1.2 #53

Closed
chovanecm opened this issue Jun 29, 2017 · 8 comments
Closed

Launching TensorBoard 1.2 #53

chovanecm opened this issue Jun 29, 2017 · 8 comments

Comments

@chovanecm
Copy link
Owner

TensorBoard 1.2 outputs something like
Starting TensorBoard b'54' at http://0.0.0.0:6006 when starting, which is not what Sacredboard expects to detect the port.

@chovanecm
Copy link
Owner Author

See the tensorboard.py.
Proposal: use the chain of responsibility design pattern to handle all possible cases

@enricoschroeder
Copy link

Hi there,
any ETA on Tensorboard 1.2/1.3 integration?
Thanks

@chovanecm
Copy link
Owner Author

Hi @enricoschroeder. Unfortunately, not. I must admit I haven't worked on Sacredboard much during the summer.
But there is a fork in @hermannsblum's repository that seems to include several patches of the part of the code.
If it works, I'd be happy to include it back to Sacredboard if the authors are willing to contribute it.

@hermannsblum
Copy link
Contributor

I myself forked from @LukasMasuch, who fixed the tensorboard issue.
I am using sacredboard for my thesis and therefore working on quite specialized improvements. If there is something more stabel coming out of this I will polish it for a PR. Until then I recommend the patch from https://github.com/LukasMasuch/sacredboard

@chovanecm
Copy link
Owner Author

The current dev version fixes the issue of starting Tensorboard. Tested on Linux (Ubuntu). However, tensorboard can be only started once as the port is currently fixed to 6006.
It can be stopped by navigating to /tensorboard/stop

We need a task manager or something in the UI to manage running tensorboards

@chovanecm
Copy link
Owner Author

Closing, further work is #93 and #92

@hermannsblum
Copy link
Contributor

@chovanecm While I still didn't have the time to organize the code properly, I am using a setup where sacredboard and tensorboard are running inside a docker container on any machine you like, and event files are loaded into the tensorboard by selecting them in sacredboard.
I think this would solve a lot of issues, but as I said I will need some time to put it into a proper setup than can be used also by other people.

@chovanecm
Copy link
Owner Author

chovanecm commented May 12, 2018

@hermannsblum That sounds interesting. I have for now at least managed to get multiple tensorboard instances running at the same time (#93).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

3 participants