-
Notifications
You must be signed in to change notification settings - Fork 448
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
Support experiments with generate_name feature #2243
Comments
That makes sense. We should get the generated name if the name is empty here: katib/sdk/python/v1beta1/kubeflow/katib/api/katib_client.py Lines 104 to 114 in fc6beec
cc: @andreyvelich |
@tenzen-y I can work on this if this feature is still needed. |
Yes, I believe that this is still needed. Feel free to open the PR. |
@tenzen-y katib/pkg/ui/v1beta1/backend.go Line 130 in fc6beec
|
@bharathk005 That backend is only for Katib UI. So, we can improve only SDK. |
@bharathk005 Could you say |
/assign |
This should be fixed with: #2272 |
/kind feature
Describe the solution you'd like
The Python katib client (again) supports experiments with the
generate_name
feature.Anything else you would like to add:
After upgrading from kubeflow-katib 0.14.0 to 0.16.0, I noticed that the
create_experiment
function inkatib_client.py
now implicitly makes the assumption that theexperiment
object contains an attribute called.metadata.name
. However, generallyname
can be missing in favour ofgenerate_name
which is then treated as a prefix which gets an additional suffix from the server. In version 0.14.0, the server return was used to access the resulting final name. At least from kubeflow-katib version 0.15.0 onwards, the server return is no longer captured and the original experiment object is used to access thename
attribute.I was wondering if the change was a conscious decision or rather a side effect that could be reverted. I am also happy to collaborate in case this is something that makes sense for katib to support.
Thanks
Thomas
Love this feature? Give it a 👍 We prioritize the features with the most 👍
The text was updated successfully, but these errors were encountered: