-
Notifications
You must be signed in to change notification settings - Fork 9
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
Unable to configure the number of GPUs in the Jupyter Notebook UI #36
Comments
This is at minimum a feature request (add this as config for the jupyter-notebooks charm), but might also be a bug upstream if it isn't reading the value properly. |
Looking back at this, this might just be an easy spawner config change. The comments aren't completely clear, but I think this might be where we can configure the gpus? I wonder if our current setup is just the default, but that we can override it here |
Hi ca-scribner, |
Hi! Screenshots from pre-release testing of 1.7 |
It is same in 1.6. Also it was mentioned you should be able to choose from 1-8 not only 1,2,4,8. It should be 1,2,3,4,5,6,7,8 as in NVIDIA GPU you can define make GPU available 1-8 any number. |
Looks like number of GPUs is also specified in frontend: |
Possible solution is to uncomment and change this line in
And change HTML in upstream:
We need someone proficient in frontend to take a look. |
Tried the change suggested in PR. The change did not solve the issue. See testing instructions in the PR. |
Hey folks, indeed it's as @i-chvets mentioned. This is a missing feature in upstream itself. Once kubeflow/kubeflow#6550 or kubeflow/kubeflow#7124 gets merged into upstream, most probably in upstream KF 1.8, then we'll also get these. |
It would be beneficial to be able to modify the Jupyter configmap to allow users to request even numbers of GPUs in the Jupyter notebook UI. This is controlled by the jupyter configmap “jupyter-ui-config-config”’s spawner_ui_config.yaml key.
Changing it manually does not seem to get overwritten by Juju.
Restarted the pod after applying the change and the change is still in place.
$ kubectl edit cm -n kubeflow jupyter-ui-config-config
$ kubectl delete pod -n kubeflow jupyter-ui-589594dd59-c6j26
The text was updated successfully, but these errors were encountered: