No check for logdir path
See original GitHub issueVersion:
TensorBoard 2.1.0
Issue:
TensorBoard doesn’t check if logdir
exists or not.
Not a huge issue, but a bit weird that there is no error message if I provide a non-existing directory. If one mistakenly provides a faulty path, they will probably only notice it when opening the link in the browser, after which they have to close the tab, go back, kill the process and restart it. It would be nice to have a warning at least.
Suggestion
Start Tensorboard as usual, but warn the user if logdir
does not exist (yet).
Issue Analytics
- State:
- Created 4 years ago
- Reactions:1
- Comments:5 (1 by maintainers)
Top Results From Across the Web
TensorBoard --logdir="path/to/log" (cannot assign to operator)
If yes, from the top directory of that folder, you open cmd (no python idle, command prompt) write tensorboard --logdir=name_of_the_folder.
Read more >logDir has default path?when I set tf.summary.FileWriter ...
TensorBoard resolves a relative logdir path against the current directory. So if you are in a directory called "/checkpoints" then running ...
Read more >Get started with TensorBoard - TensorFlow
TensorBoard is a tool for providing the measurements and visualizations needed during the machine learning workflow. It enables tracking ...
Read more >logs - AWS IoT Greengrass
--log-dir , -ld . The path to the directory to check for log files, such as /greengrass/v2 /logs . Do not use with...
Read more >Solved: Help with Changing Syslog.global.LogDir
![Change location of VMware ESXi System Logs to local or remote datastore - 2][3]; Browse datastore where we store ESXi logs to check...
Read more >
Top Related Medium Post
No results found
Top Related StackOverflow Question
No results found
Troubleshoot Live Code
Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start Free
Top Related Reddit Thread
No results found
Top Related Hackernoon Post
No results found
Top Related Tweet
No results found
Top Related Dev.to Post
No results found
Top Related Hashnode Post
No results found
Yep, that’s why we’ve historically not done this, and in fact considered it a blocker for designs that would rely on this, such as better diagnostics for accidental misuse of
--logdir_spec
.For instance, in Colab/Jupyter, one runs
%tensorboard --logdir ...
prior to starting the model training, because while the model is training you can’t execute any other Colab cells, but if you’ve already started TensorBoard then it will continue to live-update.There could be some avenue for improving the usability. Perhaps a flag or config to allow users to specify the desired behavior. What designs wouldn’t interrupt current behavior but would make lives easier?