Skip to main content

Use notebooks

Supported notebook images

Notebook instances in KakaoCloud Kubeflow provide a web-based machine learning development environment.
These notebook instances run based on Docker images, and the following images are currently supported in KakaoCloud Kubeflow.

info

The image registry endpoint is bigdata-150.kr-central-2.kcr.dev/kc-kubeflow/(image-name).
For example, to pull the jupyter-scipy:v1.8.0.py38.1a image, use bigdata-150.kr-central-2.kcr.dev/kc-kubeflow/jupyter-scipy:v1.8.0.py38.1a.

Supported notebook images

ImagePython versionPlatformFramework (version)GPU supportHDE support
kc-kubeflow/jupyter-scipy:v1.8.0.py38.1apython 3.8.10JupyterLabscipy(1.10.1)XX
kc-kubeflow/jupyter-scipy:v1.8.0.py311.1apython 3.11.9JupyterLabscipy(1.11.3)XX
kc-kubeflow/jupyter-pytorch-full:v1.8.0.py38.1apython 3.8.10JupyterLabpytorch(2.3.0)XX
kc-kubeflow/jupyter-pytorch-full:v1.8.0.py311.1apython 3.11.9JupyterLabpytorch(2.3.0)XX
kc-kubeflow/jupyter-pytorch-cuda-full:v1.8.0.py38.1apython 3.8.10JupyterLabpytorch(2.3.0)OX
kc-kubeflow/jupyter-pytorch-cuda-full:v1.8.0.py311.1apython 3.11.9JupyterLabpytorch(2.3.0)OX
kc-kubeflow/jupyter-tensorflow-full:v1.8.0.py38.1apython 3.8.10JupyterLabtensorflow(2.13.1)XX
kc-kubeflow/jupyter-tensorflow-full:v1.8.0.py311.1apython 3.11.9JupyterLabtensorflow(2.15.1)XX
kc-kubeflow/jupyter-tensorflow-cuda-full:v1.8.0.py38.1apython 3.8.10JupyterLabtensorflow(2.13.1)OX
kc-kubeflow/jupyter-tensorflow-cuda-full:v1.8.0.py311.1apython 3.11.9JupyterLabtensorflow(2.15.1)OX
kc-kubeflow/jupyter-pyspark-pytorch:v1.8.0.py38.1apython 3.8.10JupyterLabpytorch(2.3.0)XO
kc-kubeflow/jupyter-pyspark-pytorch:v1.8.0.py311.1apython 3.11.9JupyterLabpytorch(2.3.0)XO
kc-kubeflow/jupyter-pyspark-pytorch-cuda:v1.8.0.py38.1apython 3.8.10JupyterLabpytorch(2.3.0)OO
kc-kubeflow/jupyter-pyspark-pytorch-cuda:v1.8.0.py311.1apython 3.11.9JupyterLabpytorch(2.3.0)OO
kc-kubeflow/jupyter-pyspark-tensorflow:v1.8.0.py38.1apython 3.8.10JupyterLabtensorflow(2.13.1)XO
kc-kubeflow/jupyter-pyspark-tensorflow:v1.8.0.py311.1apython 3.11.9JupyterLabtensorflow(2.15.1)XO
kc-kubeflow/jupyter-pyspark-tensorflow-cuda:v1.8.0.py38.1apython 3.8.10JupyterLabtensorflow(2.13.1)OO
kc-kubeflow/jupyter-pyspark-tensorflow-cuda:v1.8.0.py311.1apython 3.11.9JupyterLabtensorflow(2.15.1)OO
kc-kubeflow/codeserver-python:v1.8.0.py38.1apython 3.8.10CodeServerXX
kc-kubeflow/codeserver-python:v1.8.0.py311.1apython 3.11.9CodeServerXX

Create notebook instance

To build a computing environment for running machine learning code and processing data, create a notebook instance with your desired specifications.

  1. Access the Kubeflow dashboard.

  2. Click the Notebooks tab on the left panel.

  3. Click the [New Notebook] button at the top right.

    Image. Accessing the Notebooks tab in Kubeflow dashboard
    Accessing the Notebooks tab in Kubeflow dashboard

  4. On the New Notebook screen, fill in the required information and click the [LAUNCH] button to create the notebook instance.

    • If you want to create a notebook using a specific image, refer to the guide below.
    ItemFieldDescription
    NameNameIdentifier used in the Kubeflow dashboard
    Platform typeWeb development platform type
    Options: JupyterLab, VisualStudio Code, RStudio
    *VisualStudio Code and RStudio may not work properly without a connected domain.
    Custom notebookImageSelect a supported KakaoCloud Kubeflow notebook image
    Custom imageEnter a custom image by checking the box
    - Example: bigdata-150.kr-central-2.kcr.dev/kc-kubeflow/jupyter-pytorch-full:v1.0.1.py38
    Image pull policySet the policy for pulling the image when starting the notebook
    CPU / RAMMinimum CPUMinimum CPU to allocate to the instance
    - Labeled as Requested CPUs in version 1.6
    Minimum memory GiMinimum memory to allocate
    - Labeled as Requested memory in Gi in version 1.6
    Maximum CPUMaximum CPU to allocate
    - Labeled as CPU limit in version 1.6
    Maximum memory GiMaximum memory to allocate
    - Labeled as Memory limit in Gi in version 1.6
    GPUsNumber of GPUsNumber of GPU resources to use
    GPU vendorAvailable GPU instance types
    Workspace volumeVolume to mount to the notebook’s home directory
    - Can create new or use an existing volume
    Data volumeAdditional volumes to mount
    - Define mount paths and choose/create volumes
    ConfigurationsRegister PodDefault resources in the current namespace
    Affinity / TolerationsAffinity configSpecify the node on which to run the notebook
    Tolerations groupSet tolerations
    - Toleration list is defined in the ConfigMap within KE
    Miscellaneous settingsEnable shared memoryEnable shared memory support (e.g., for Torch)
    - Mounts emptyDir to /dev/shm
caution

Notebook instances will not be created if Minimum CPU or Memory Gi exceeds the available resources of the selected node pool in Affinity Config.

Create CPU-based notebook

To create a notebook using a CPU-based image:

  1. Access the Kubeflow dashboard and click the Notebooks tab, then click the [New Notebook] button.

  2. On the New Notebook screen, enter the required info and click [LAUNCH].

    ItemFieldDescription
    NameNameIdentifier used in the Kubeflow dashboard
    Platform typeJupyterLab, VisualStudio Code, or RStudio
    Custom notebookImageSelect an image matching the Kubeflow version
    Custom imageEnter custom image if needed
    Image pull policySet image pull policy
    CPU / RAMMinimum CPU2
    Minimum memory Gi8
    Maximum CPU-
    Maximum memory Gi-
    GPUsNumber of GPUsNone
    GPU vendor-
    Workspace volumeMount volume to home directory
    Data volumeRegister and configure additional mount volumes
    ConfigurationsRegister PodDefault resource
    Affinity / TolerationsAffinity configSelect CPU-type node
    Tolerations groupNone
    Miscellaneous settingsEnable shared memoryMount emptyDir to /dev/shm

Create GPU-based notebook

To create a notebook using a GPU-based image:

  1. Access the Kubeflow dashboard and click the Notebooks tab, then click the [New Notebook] button.

  2. On the New Notebook screen, enter the required info and click [LAUNCH].

    ItemFieldDescription
    NameNameIdentifier used in the Kubeflow dashboard
    Platform typeJupyterLab, VisualStudio Code, or RStudio
    Custom notebookImageSelect an image matching the Kubeflow version
    Custom imageEnter custom image if needed
    Image pull policySet image pull policy
    CPU / RAMMinimum CPU2
    Minimum memory Gi8
    Maximum CPU-
    Maximum memory Gi-
    GPUsNumber of GPUs1
    GPU vendorNVIDIA MIG - 1g.10gb
    Workspace volumeMount volume to home directory
    Data volumeRegister and configure additional mount volumes
    ConfigurationsRegister PodDefault resource
    Affinity / TolerationsAffinity configSelect GPU-type node
    Tolerations groupNone
    Miscellaneous settingsEnable shared memoryMount emptyDir to /dev/shm

Access notebook instance

  1. Access the Kubeflow dashboard.

  2. Click the Notebooks tab.

  3. In the Notebooks list, click [CONNECT] next to the desired instance.

    Image. Connect to notebook
    Connect to notebook

  4. A new tab opens with the Jupyter Notebook interface.

    • You can write code and run ML models there.

    Image. Jupyter Notebook interface
    Jupyter Notebook interface

Stop notebook instance

Notebook instances can be stopped to save resources, improve security, reduce cost, and manage environments.

  1. Access the Kubeflow dashboard.

  2. Click the Notebooks tab.

    Image. Stop notebook
    Stop notebook

  3. Click the ◼ (Stop) button for the notebook.

    • In the confirmation modal, click [STOP].

    Image. Stop notebook confirmation
    Stop notebook confirmation

  4. Check that the status of the notebook shows as stopped.

    Image. Notebook stopped
    Notebook stopped

Delete notebook instance

Notebook instances are deleted to reclaim resources, improve security, reduce cost, and maintain clean environments.

  1. Access the Kubeflow dashboard.

  2. Click the Notebooks tab.

    Image. Delete notebook
    Delete notebook

  3. Click the trash icon for the notebook you want to delete.

    • In the popup, click [DELETE].

    Image. Delete notebook confirmation
    Delete notebook confirmation

  4. Verify that the notebook is no longer listed.

    Image. Notebook deleted
    Notebook deleted

caution
  • Deleted notebook instance names cannot be reused.
  • If reuse is needed, delete the {notebook-name}_volume PVC in the KakaoCloud Kubernetes Engine cluster before recreating.
info