Skip to content

アプリケーション内で をクリックすると、お使いのDataRobotバージョンに関する全プラットフォームドキュメントにアクセスできます。

SAP AI Coreでのスコアリングコードの自動デプロイと置換

本機能の提供について

Automated deployment and replacement of Scoring Code in SAP AI Core is a premium feature, off by default. この機能を有効にする方法については、DataRobotの担当者または管理者にお問い合わせください。

Feature flag: Enable the Automated Deployment and Replacement of Scoring Code in SAP AI Core (Premium feature)

DataRobotが管理するSAP AI Core予測環境を作成して、SAP AI CoreでDataRobotスコアリングコードをデプロイします。 DataRobot管理を有効にすると、外部SAP AI Coreデプロイは、自動スコアリングコード置換を含むMLOps管理にアクセスできます。

Service health information for external models and monitoring jobs

Service health information such as latency, throughput, and error rate is unavailable for external, agent-monitored deployments or when predictions are uploaded through a prediction monitoring job.

SAP AI Core予測環境の作成

SAP AI Coreにモデルをデプロイするには、まずカスタムSAP AI Core予測環境を作成します。

  1. コンソールで、予測環境+ 予測環境を追加の順にクリックします。

  2. 予測環境を追加ダイアログボックスで、予測環境設定を行います。

    • 予測環境の説明的な名前とオプションの説明を入力します。

    • プラットフォームドロップダウンリストからSAP AI Coreを選択します。

    • DataRobotが管理設定を有効にすると、この予測環境が管理エージェントを介してDataRobotスコアリングコードモデルを自動的にパッケージ化してデプロイできます。

    The Supported model formats settings are automatically set to DataRobot and DataRobot Scoring Code only and can't be changed, as this is the only model format supported by DataRobot managed SAP AI Core. In addition, SAP AI Core prediction environments do not support time series models.

  3. In the Management settings, select the related SAP credentials and SAP resource group.

    SAP Oauth credentials required

    DataRobot management of Scoring Code in SAP AI Core requires existing SAP Credentials. If you don't have existing credentials, the No SAP credentials found alert appears, directing you to Manage credentials to create SAP AI Core credentials.

    To create the required credentials, for Credential type, select SAP OAuth. Then, enter a SAP API URL, Auth URL, Client ID, Client secret, and a Display name. 資格情報を検証して保存するには、保存してサインインをクリックします。

  4. In the Monitoring settings, click Enable monitoring and optionally, define Environment variables.

  5. 環境設定を行った後、環境を追加をクリックします。

    The SAP AI Core environment is now available from the Prediction environments page.

Deploy a model to the SAP AI Core prediction environment

SAP AI Core予測環境を作成したら、その環境にモデルをデプロイできます。

  1. On the Registry > Model directory page, in the table of registered models, click the registered model containing the version you want to deploy, opening the list of versions.

    モデルサポート

    SAP AI Core prediction environments do not support time series models or models without Scoring Code support.

  2. From the list of versions, click the Scoring Code enabled version you want to deploy, opening the registered model version panel.

  3. In the upper-right corner of any tab in the registered model version panel, click Deploy.

  4. In the Prediction history and service health settings, under Choose prediction environment, click Change.

  5. In the Select prediction environment panel, click SAP AI Core, and then click the prediction environment you want to deploy to.

  6. With a SAP AI Core environment selected, under SAP resource plan, select a plan based on the anticipated CPU and memory usage of your prediction workloads. For more information on these resource plans, see the SAP AI Core documentation:

  7. (Optional) Open the Advanced environment settings and define additional Environment key-value pairs to provide extra parameters to the SAP AI Core deployment interface.

  8. Configure the remaining deployment settings, and then click Deploy model.

While the deployment is Launching, you can monitor the status events on the deployment's Monitoring > Service health tab under Recent activity > Agent activity

SAP AI Coreで予測を行う

After you deploy a model to an SAP AI Core prediction environment, you can use the code snippet from the Predictions > Portable predictions tab to score data.

Prediction request payload limit

The maximum prediction request payload is 1MB (approximately 2000 rows) per request. The provided code snippet splits the prediction payload into multiple requests.

Before you run the code snippet, you must export environment variables containing the secrets associated with the Service Key used SAP AI Core:

環境変数 説明
SAP_AI_API_URL The URL of the SAP AI Core service.
SAP_AI_AUTH_URL The URL used for authentication with SAP AI Core.
SAP_CLIENT_ID The client ID associated with your SAP AI Core Service Key.
SAP_CLIENT_SECRET The client secret associated with your SAP AI Core Service Key.

備考

These are the same fields provided when creating a SAP OAuth credential.

機能に関する注意事項

  • Only Scoring Code JAR-enabled models are supported.

  • Custom models, LLMs, and time series models are not supported.

  • Challenger models and model replacement are not supported (challenger prediction servers can't be set to an external or serverless prediction environment).

  • Batch monitoring is not supported.

  • Only CSV files are supported for predictions. XLSX files are not supported by the code snippet.

  • The maximum prediction request payload is 1MB (approximately 2000 rows) per request. The code snippet splits the prediction payload into multiple requests.


更新しました September 18, 2024