Global AI and Data Science

Global AI & Data Science

Train, tune and distribute models with generative AI and machine learning capabilities

 View Only
  • 1.  Train AutoAI and reference model run error

    Posted Fri April 21, 2023 10:15 AM
      |   view attached

    hi there,
    I am a new learner of  machine learning.
    when I built a project of Train AutoAI and reference model ,and run it.
    I always got the following errors:

    translating TaskSpec to Pod: serviceaccounts "pipeline-runner" not found. Maybe invalid TaskSpec

    I really don't know why , anyone could tell me? . thank you very much.



    ------------------------------
    wang xiangtao
    ------------------------------

    #AIandDSSkills


  • 2.  RE: Train AutoAI and reference model run error

    Posted Mon April 24, 2023 08:06 AM

    Hi, it seems you cannot even launch the pipeline. Can you check if your project has associated a Watson Machine Learning isntance? Go to your project , Manage=> Service & Integrations, here you should add a valid Watson Machine Learning instance



    ------------------------------
    Daniel Lopez Sainz
    ------------------------------



  • 3.  RE: Train AutoAI and reference model run error

    Posted Thu May 11, 2023 09:15 PM

    Hi, thanks you for your answer .  I  associated a Watson Machine Learine instance. So It seems have other reason.



    ------------------------------
    wang xiangtao
    ------------------------------



  • 4.  RE: Train AutoAI and reference model run error

    Posted Thu April 27, 2023 09:07 PM

    Hello,

    It seems like you are encountering an issue related to the service account "pipeline-runner" not being found in your project. This might be due to a missing or incorrectly configured service account.

    Here are a few suggestions to help you resolve this issue:

      1. Ensure that the service account "pipeline-runner" has been created in the appropriate namespace. You can check this using the following command:
    kubectl get serviceaccounts -n <NAMESPACE>

    Replace <NAMESPACE> with the namespace you are working in. If "pipeline-runner" is not listed, you may need to create it.

      1. To create the service account, you can apply the following YAML file:
    apiVersion: v1
    kind: ServiceAccount
    metadata:
      name: pipeline-runner
      namespace: <NAMESPACE>

    Replace <NAMESPACE> with your desired namespace and save the file as pipeline-runner.yaml. Apply the file using the command:

    kubectl apply -f pipeline-runner.yaml
    1. Ensure that the required permissions are granted to the "pipeline-runner" service account. You may need to create a ClusterRoleBinding or RoleBinding depending on your requirements.
    2. Double-check your TaskSpec definition and ensure that it is configured correctly. Make sure that the service account name specified in the TaskSpec matches the one you created (i.e., "pipeline-runner").

    If you follow these steps and still encounter issues, please provide more information about your project setup, including the version of the software being used, so that we can further assist you.

    Best of luck with your machine learning project!



    ------------------------------
    Scott Dunn
    ------------------------------



  • 5.  RE: Train AutoAI and reference model run error

    Posted Fri May 12, 2023 12:14 AM

    Hi , Thanks for your answer.  I retried several days ago , And it has been successfully executed.
    But I don't know how to execute your command in Watson Machine Learning instance. 
    I use the instance of lite level, Maybe I have no permission to do that.



    ------------------------------
    wang xiangtao
    ------------------------------



  • 6.  RE: Train AutoAI and reference model run error

    Posted Wed May 17, 2023 09:24 AM

    Following up! Did you got the solution for the porblem or not .. Can you please update us. 

    Thank you



    ------------------------------
    Marcel Kimball
    ------------------------------