-
```
> kubectl get secret user-gcp-sa -o yaml
Fetching cluster endpoint and auth data.
kubeconfig entry generated for cluster-1.
apiVersion: v1
data:
application_default_credentials.json: ""
…
-
Since we decided to remove service account credential from our MKP deploy page, the user will need to figure out how to correctly apply SA credential after deploying KFP themselves. A clear doc/guidel…
-
This can be reproduced by running the latest prebuilt tfx samples and open the tensorboard in the training step through GCP marketplace.
-
Currently, it is possible for a user to upload a notebook sample with hard-coded arbitrary experiment name to run their pipeline. However, our sample test infra adopts the convention that the experime…
-
When running a pipeline with a ParallelFor condition over a list of dictionaries, camelCase and snake_case inputs are altered.
This won't work:
with kfp.dsl.ParallelFor([{"helloWorld": "bl…
-
**What happened:**
When calling ParallelFor on the output of a lightweight python component, I get an error saying that the task is not defined.
```py
@dsl.pipeline(
name="Parallel Pipelin…
-
Context #2342
Working on a fix.
-
**What happened:**
I continuously, but somewhat sporadically, experiencing extremely long loading times for (at least) pipelines and experiments in the UI. When issue occurs the loading time can vary…
-
**What happened:**
by default, we use kfp.client or dsl.compile, generated workflow yaml is without **metadata.namespace**. due to this pipeline always runs in default kubeflow namespace.
**What …
-
The current behavior is:
The output argument is injected for both normal python sample and notebook sample when it is specified in the config yaml.
The experiment_name is only injected for notebook…