Skip to content

PoC using the `nice` *nix command for machine learning, to monitor server-wide resource utilization

Notifications You must be signed in to change notification settings

jake-mason/nice-ml

Repository files navigation

nice-ml

PoC using the nice command in Linux

Getting started

First, use virtualenv to build the Python virtual environment containing a few necessary third-party packages:

$ ./create_virtual_environment.sh

If the environment isn't automatically accessed, run source venv/bin/activate. (venv) should appear at the left of your console.

Next, open a screen session and run ./runner.sh. You'll be emailed (at least, I'll be emailed, unless you change the email in runner.sh) once the job completes.

Note: although machine learning training workflows are very resource-intensive, the frequency with which these jobs run is completely dependent on your use case. Typical retraining schedules are weekly or monhtly, so whatever level of resource utilization accompanies your job, know that job won't be running all day, every day.

About

PoC using the `nice` *nix command for machine learning, to monitor server-wide resource utilization

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published