Set up Jupyter Notebook environment using Kubeflow

Learn how to perform data analysis and model training using Jupyter Notebook, and optimize machine learning workflows using various features.

Scenario overview

In this scenario, you will be guided through the process of accessing the dashboard and then creating a Jupyter Notebook instance. The main contents of this scenario are as follows.

• Perform data analysis and model training after creating a Jupyter Notebook
• Build machine learning workflows using various components and features

Step 1. Access dashboard

Access the dashboard. Users can manage various resources and configure the Jupyter Notebook environment through the dashboard.

Step 2. Create Jupyter Notebook

Through the dashboard, users can create a Jupyter Notebook instance. In this step, you select the notebook specifications and proceed with the required settings.

1. Click Notebooks in the left tab of the dashboard.

2. Move to the Notebooks page and click the [+ New Notebook] button located at the top right. Refer to the information below and create a new notebook.

CPU-based notebook image

1. Refer to the following information to create a CPU-based notebook image.

   Item	Category	Description
   Name	Name	Used to identify the notebook instance in the Kubeflow dashboard
   	Namespace	Kubernetes namespace where the notebook instance is created
   Docker Image	Image	Select the Docker image to use - Check supported notebook images
   CPU / RAM	Requested CPUs	2 - Number of CPU cores allocated to the notebook instance
   	Requested memory in Gi	8 - Memory resource unit (GiB) allocated to the notebook instance
   Tolerations	Tolerations Group	None - Allow taints of specific nodes

2. Enter the notebook information to be created. Refer to the table below for example values.

   Item	Value
   Name	handson
   Image	kc-kubeflow/jupyter-pytorch-cuda-full:v1.0.1.py36
   Requested CPUs	2
   Requested memory in GiB	8

3. Click the [LAUNCH] button to create the notebook.

info

You can create a Jupyter Notebook using a Custom Image when selecting the Docker Image.

Step 4. Access Jupyter Notebook

Access the created Jupyter Notebook instance to proceed with an actual machine learning project.

1. Click the [CONNECT] button for the created notebook instance to access it.

   

2. In the Notebook, select the Python3 kernel, enter a simple example code, and check the execution result.