Service development

更新时间:
复制 MD 格式

AI Search Open Platform integrates PAI DSW, a zero-configuration notebook environment built on JupyterLab, so you can code, debug, and run Python services end-to-end without leaving the platform.

Billing

After you start an instance, you are charged on a pay-as-you-go basis. The fee is calculated using the following formula: CU price × CUs per instance type × Number of instances.

The following table describes the billing rules.

Instance type

CU price (CNY/hour)

Number of CUs consumed per machine

Price per machine (CNY/hour)

gpu.t4.16g.x1

1.07

16.07

17.195

ops.basic1.gi.large

0.61

0.65

For example, if you select one ops.basic1.gi.large instance, the fee is calculated as follows: 1.07 × 0.61 × 1 = 0.65 CNY/hour.

Create instance

  1. In AI Search Open Platform, navigate to Model Service > Service Development, and then click Create Development Instance.

  2. Enter an Instance Name and Description, select a Resource Type, and then submit. The system begins deploying the instance.

    Instance type

    CPU cores and memory

    GPU and video memory

    gpu.t4.16g.x1

    8 cores, 31 GB

    1 × NVIDIA T4 16 GB

    ops.basic1.gi.large

    2 cores, 8 GB

    N/A

    The Instance Name must start with a letter, contain only letters, digits, and underscores (_), and be no more than 30 characters long. The Description must be no more than 20 characters long. The default Billing Method is pay-as-you-go. Instance Name and Resource Specification are required.

    An instance can have the following statuses:

    • Deploying: The instance is being provisioned. You can delete it during this state.

    • Running: The instance is active. You can click Go to Development to open the notebook development environment, click Stop to stop the instance, or click image>Delete to delete the instance.

    • Stopped: The instance is not running. You can click Start to restart the instance, or click image>Delete to delete it.

    • Frozen: The instance is frozen due to an overdue payment on your account.

Service development

When an instance is Running, you can access its notebook environment in any of the following ways:

  • In the left-side navigation pane, choose Model Service > Service Development. This page lists your development instances with details such as name, creation time, status, and description. In the Actions column, click Go to Development to open the notebook environment.

  • In the left-side navigation pane, click Scenario Center. On the RAG Scenario - Online Q&A with Knowledge Base page, click Develop in the upper-right corner. In the Select Instance dialog box, select an instance such as doc_01_test, and click Go to Development.

  • In the left-side navigation pane, click Experience Center. Select a target service, then in the code panel on the right, click the Sample Code tab. Click >>Develop in the upper-right corner. In the Select Instance dialog box, select an instance and click Go to Development. If no instances are available, create one under Service Development first.