You must create Data Science Workshop (DSW) instances before you use DSW to build Notebook models. This topic describes how to create a DSW instance.

Prerequisites

If you are a first-time user of DSW, you must first assign service-linked roles to DSW so that DSW can access the required resources. For more information, see Assign service-linked roles.

If you want to create a DSW instance as a RAM user, you must first use your Alibaba Cloud account to authorize the RAM user. For more information, see Authorize RAM users.

Procedure

  1. Log on to the Machine Learning Platform for AI console.
  2. In the left-side navigation pane, choose Model Training > DSW-Notebook Service.
  3. In the upper-left corner of the page, select the region where you want to create the instance.
  4. On the Notebook Modeling page, click Create Instance.
  5. In the Configure Instance step, set the parameters that are described in the following table.
    Parameter Description
    Instance Name The name cannot exceed 27 characters in length. It can contain only letters, digits, and underscores (_).
    Instance Version The following editions are supported:
    • DSW Individual Edition: This edition is upgraded from DSW 2.0. It is developed based on cloud-native technologies of Alibaba Cloud, such as Docker and Kubernetes. It provides open and AI-assisted development environments for you to train models with high elasticity.
    • DSW On-sale Edition: This edition is upgraded from DSW 1.0. It is developed based on the Apsara big data platform of Alibaba Cloud and greatly reduces the costs of running DSW instances. However, this edition does not support Internet access, root permissions, or sudo operations. Purchase with caution.
    Region & Zone Instances in different regions cannot communicate with each other. Select a region that is close to your clients to reduce network latency and improve connection quality.
    Billing Method A DSW Individual Edition instance supports only the pay-as-you-go billing method. A DSW On-sale Edition instance supports only the subscription billing method.
    Network This parameter is available only to DSW On-sale Edition instances. The default setting is Classic Network, and cannot be changed.
    Instance This parameter is available only to DSW Individual Edition instances. You can create a CPU- or GPU-accelerated instance. For more information about instance specifications, see Individual Edition.
    Resource Type This parameter is available only to DSW On-sale Edition instances. If the number of remaining GPU cards is insufficient for your business needs, go to the buy page to purchase GPU cards.
    Storage The instance comes with a system disk storage as a temporary storage. After the instance is stopped or deleted, this storage is cleared. If a permanent storage is required, you must mount a created Apsara File Storage NAS file system to the instance. For more information about how to create an Apsara File Storage NAS file system, see Manage file systems.
    Note After an Apsara File Storage NAS file system is mounted, the instance uses this permanent storage, instead of the temporary storage, to store data.
    Select Image Multiple Python, TensorFlow, and PyTorch versions of images are supported. For more information about supported images, see Individual Edition.
    VPC This parameter is available only to DSW Individual Edition instances. You can select a VPC to deploy the DSW instance. If you set the VPC parameter, you must also set the vSwitch and Security Group parameters.

    You can select an existing VPC to deploy the DSW instance. You can also click Create VPC next to VPC and create a VPC to deploy the DSW instance.

    vSwitch If you set the VPC parameter, you must also set the vSwitch and Security Group parameters.

    You can select an existing vSwitch and an existing security group to deploy the DSW instance. You can also click Create vSwitch and Create Security Group and create a vSwitch and a security group to deploy the DSW instance.

    Security Group
  6. Click Confirm Order.
  7. After you confirm the order, select Machine Learning DSW Terms of Service and click Create Instance.