Background

• You will be charged for creating and using the notebook instances. The fees vary according to your selected resources. For details about pricing, see Product Pricing Details.
• You will be charged as long as your notebook instance is in the Running status. We recommend you to stop the notebook instance when you no longer need it to avoid unnecessary fees. Alternatively, you can enable the auto stop function when creating a notebook instance so that the notebook instance can automatically stop at the specified time without incurring unnecessary fees.
• Only notebook instances in the Running state can be started.
• A maximum of 10 notebook instances can be created for an account. By default, the notebook instances created by the current user are displayed. If Display Only My Instances is disabled, all notebook instances created by the current account and its IAM users are displayed.
• ModelArts supports multiple AI engines. When creating a notebook instance, select the Python2 or Python3 environment. After the notebook instance is created, access it and create environments with desired AI engines on the Jupyter page.
• If OBS storage is used, ensure that the OBS directory you use and ModelArts are in the same region.
• Before creating a notebook instance, learn about the AI engines supported by ModelArts and their versions.

Creating a Notebook Instance

1. Log in to the ModelArts management console. In the left navigation pane, choose DevEnviron > Notebooks to switch to the Notebooks page.
2. Click Create. On the displayed page, set the required parameters.
   1. Enter the basic information about the notebook instance, including the name, description, and whether to automatically stop the notebook instance. For details about the parameters, see Table 1.
      Figure 1 Basic information about a notebook instance
      

      Table 1 Parameters of basic information

      Parameter	Description
      Billing Mode	Select Pay-per-use. Currently, only pay-per-use billing is supported.
      Name	Name of a notebook instance, which contains a maximum of 64 characters. Only digits, uppercase and lowercase letters, underscores (_), and hyphens (-) are allowed. This parameter is mandatory.
      Description	Brief description of a notebook instance.
      Auto Stop	This function is enabled by default. The default value is 1 hour later, indicating that the notebook instance automatically stops after running for 1 hour and its billing stops. The options are 1 hour later, 2 hours later, 4 hours later, 6 hours later, and Custom. If you select Custom, you can specify any integer from 1 to 24 hours.

   2. Set notebook parameters, such as the work environment and instance flavor. For details, see Table 2.
      Figure 2 Notebook instance parameters
      

      Table 2 Notebook instance parameters

      Parameter	Description
      Work Environment	Only public images, that is, the AI frameworks built in ModelArts are supported. All supported AI engines can be used in the same notebook instance. Different engines can be switched quickly and conveniently, and run in independent development environments. After the notebook instance is created, go to the Jupyter page to create the development environment with the desired AI engine. The AI engine varies depending on the operating environment. For details, see Supported AI Engines. Multi-Engine 1.0 (Python 3, Recommended) is recommended. If you need to use TensorFlow 2.X, PyTorch 1.4.0, or an AI framework with the R language, you are advised to use Multi-Engine 2.0 (Python3). If you select this work environment, you are advised to select a paid flavor because the free flavor is unavailable. NOTE: ModelArts also supports the Keras engine. You can select Multi-Engine 1.0 (Python 3, Recommended) or Multi-Engine 1.0 (Python 2). For details, see Does ModelArts Support the Keras Engine?
      Resource Pool	Select Public resource pools or Dedicated resource pools. For details about the dedicated resource pools and how to buy them, see Resource Pools. You do not need to purchase the public resource pool. The public resource pool is available immediately after being provisioned and is charged based on the running duration of your notebook instances. Dedicated resource pools are queue-free. When a large number of users use the public resource pool, they may queue for resources. You are advised to purchase a dedicated resource pool to improve development efficiency.
      Type	The CPU, GPU, and Ascend types are supported. The GPU type delivers better performance, but is more costly than the CPU type. The Ascend type is an OBT resource. Apply for the Ascend 910 OBT in advance.
      Instance Flavor	A flavor needs to be selected only when you select Public resource pools. The instance flavors vary with the Type setting. If you select CPU for Type, available options include [Free] CPU, 2 vCPUs | 8 GiB, and 8 vCPUs | 32 GiB. If you select GPU for Type, available options include [Free] GPU and GPU: 1 x v100NV32 CPU: 8 vCPUs | 64 GiB . If you select GPU for Type, Ascend: 1 x Ascend 910 CPU: 24 vCPUs | 96 GiB is available. If you select free specifications, read the prompt carefully and select I have read and agree to the above.
      Storage	The EVS and OBS options are available. Selecting EVS Set Disk Space based on the actual usage. The default value of Disk Space is 5 GB. ModelArts provides 5 GB disk space for you to use for free. If the disk space exceeds 5 GB, the additional space is billed by GB according to pricing of ultra-high I/O disks. The value of Disk Space ranges from 5 to 4096. If you select this storage mode, all read and write operations on files on the notebook instances take effect on the data stored in your notebook instances. There is no data loss after you restart notebook instances. Selecting OBS Click Select next to the Storage Path text box to set the OBS path for storing notebook instance data. If you want to use existing files or data, upload the files or data to the corresponding OBS path in advance. Storage Path must be set to a specific directory in an OBS bucket rather than the root directory of the OBS bucket. If you select this storage mode, all read and write operations on files on the notebook instances take effect on the data stored in your selected OBS path. To synchronize data in a file stored in OBS to a notebook instance, select the file and click Sync OBS. For details, see Synchronizing Files with OBS. There is no data loss after you restart this notebook instance.
      Git Repository	This parameter is available only when EVS is selected for Storage. After this function is enabled, you can create a notebook instance with a Git repository. The system automatically synchronizes the code repository from GitHub. For details, see Creating a Notebook Instance with a Git Repository.

3. Click Next.
4. After confirming the parameter configurations, click Submit.

   Switch to the notebook instance list. Status of the notebook instance being created is Starting. If Status of the notebook instance changes to Running, the notebook instance has been created.