Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.
And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.
System metrics: customize logging frequency
See original GitHub issuewandb --version && python --version && uname
- Weights and Biases version: version 0.9.1
- Python version: Python 3.7.4
- Operating System: MAC OS
Description
Hi there! From the documentation, I find that default metrics will be averaged every 30 seconds. I want to ask if we have any approach to change the time interval. I hope that users can have the option to choose by themselves. For example, if I have a program that runs for about 3 minutes, we might want to record system metrics every 5 or 10 seconds. If not, is there any way to add some code to make a control outside of wandb package. Thank you!
Issue Analytics
- State:
- Created 3 years ago
- Comments:8 (3 by maintainers)
Top Results From Across the Web
Top Related Medium Post
Top Related StackOverflow Question
Troubleshoot Live Code
Top Related Reddit Thread
Top Related Hackernoon Post
Top Related Tweet
Top Related Dev.to Post
Top Related Hashnode Post
@loodvn if you’re using PyTorch, using the profiler is probably your best bet for debugging GPU memory issues but I can file a ticket for adding some of these more basic tools to the UI
Hey @coolalexzb we’re currently not exposing these dynamic settings in the UI. The relevant code is here: https://github.com/wandb/client/blob/master/wandb/stats.py#L81 As a quick workaround you could override this method. In the next release of our python library we plan to expose this in the UI, this release is planned for late July.