kale
kale copied to clipboard
kubeflow-kale
Kubeflow’s superfood for Data Scientists
When Kale is installed on Kubeflow Notebook Server, the RPC error shown below repeatedly and frequently (every few minutes) brings up a pop up box with a message. Kale is...
Quick upgrade to jupyterlab 3 by running the script. Not yet manually tested. A build workflow is added to build the extension as a python package.
``` python: 3.8.8 pip: 21.3.1 kubeflow-kale: 0.7.0 ``` Aiming to use kale sdk to compile (and run) pipeline in on-prem kubeflow environment as per documentation https://docs.arrikto.com/release-1.4/user/kale/sdk/pipelines.html#procedure example `kale_sdk.py` ``` from...
``` python: 3.8.8 pip: 21.3.1 kubeflow-kale: 0.7.0 ``` Currently running pipeline with `-K` using SDK fails due to dependency on ROK. ``` python kale_sdk.py -K 2022-01-20 10:36:54 Kale rokutils:134 [INFO]...
``` python: 3.8.8 pip: 21.3.1 kubeflow-kale: 0.7.0 ``` Currently it seems it's not supported to specify custom image for the step using @step decorator. Desired functionality would be to specify...
When designing pipelines via the kfp SDK it's possible to disable the caching for a specific pipeline step (https://www.kubeflow.org/docs/components/pipelines/overview/caching/#managing-caching-staleness). Is this possible to be done via KF Kale in jupyter...
We are trying to get Katib work with Kale in a Kubeflow 1.4 environment. One blocker has been, that the standard trial image `gcr.io/arrikto/katib-kfp-trial@sha256:5cde449d6fd6d22d3b2da7b791c777a3f9594557efbe70cf5e5a016dabf20de1` carries on outdated `kfp` SDK version....
This is basically a x-post of a Katib issue I created: https://github.com/kubeflow/katib/issues/1454#issuecomment-791474023. I would like to remove some key annotations from a Katib Experiment Job template, without having to modify...
_Description of the problem:_ #Let the random-forest model be: rf_model from kale.common.serveutils import serve kfserver = serve(rf_model) #model is deployed #prepare data for prediction data = [row.tolist() for _, row...
i have created a kubeflow pipeline but it is not starting because pvc is not created but when i create volume and pvc claims manually that component started worked but...
