이 페이지에서는 Vertex AI Pipelines에서 ML 파이프라인을 정의하고 실행하기 위해 사용할 수 있는 인터페이스를 보여줍니다.
파이프라인을 정의하는 인터페이스
Vertex AI Pipelines는 Kubeflow Pipelines(KFP) SDK 또는 TensorFlow Extended(TFX) SDK를 사용하여 정의된 ML 파이프라인을 지원합니다.
Kubeflow Pipelines(KFP) SDK
대규모 정형 데이터 또는 텍스트 데이터를 처리하기 위해 TensorFlow Extended를 사용할 필요가 없는 모든 사용 사례에 KFP를 사용하세요.
Vertex AI Pipelines는 KFP SDK v1.8 이상을 지원합니다. 하지만 Vertex AI Pipelines 문서의 코드 샘플을 사용하려면 KFP SDK v2를 사용하세요.
KFP SDK를 사용하면 커스텀 구성요소를 빌드하거나Google Cloud 파이프라인 구성요소와 같은 사전 빌드된 구성요소를 재사용하여 ML 워크플로를 정의할 수 있습니다. Google Cloud 파이프라인 구성요소를 사용하면 ML 파이프라인에서 AutoML과 같은 Vertex AI 서비스를 쉽게 사용할 수 있습니다. Vertex AI Pipelines는 Google Cloud 파이프라인 구성요소 SDK v2 이상을 지원합니다.Google Cloud 파이프라인 구성요소에 관한 자세한 내용은 Google Cloud 파이프라인 구성요소 소개를 참고하세요.
Kubeflow Pipelines를 사용하여 파이프라인을 빌드하는 방법은 파이프라인 빌드를 참조하세요. Kubeflow Pipelines에 대해 자세히 알아보려면 Kubeflow Pipelines 문서를 참조하세요.
TensorFlow Extended(TFX) SDK
ML 워크플로에서 TensorFlow Extended를 사용하여 테라바이트 규모의 정형 데이터 또는 텍스트 데이터를 처리하는 경우 TFX를 사용하세요. Vertex AI Pipelines는 TFX SDK v0.30.0 이상을 지원합니다.
Google Cloud 콘솔은 파이프라인 실행을 검토하고 모니터링하는 데 권장되는 방법입니다. Google Cloud 콘솔을 사용하여 파이프라인 실행 만들기, 삭제, 클론, 템플릿 갤러리에 액세스, 파이프라인 실행의 청구 라벨 가져오기와 같은 다른 작업을 수행할 수도 있습니다.
[[["이해하기 쉬움","easyToUnderstand","thumb-up"],["문제가 해결됨","solvedMyProblem","thumb-up"],["기타","otherUp","thumb-up"]],[["이해하기 어려움","hardToUnderstand","thumb-down"],["잘못된 정보 또는 샘플 코드","incorrectInformationOrSampleCode","thumb-down"],["필요한 정보/샘플이 없음","missingTheInformationSamplesINeed","thumb-down"],["번역 문제","translationIssue","thumb-down"],["기타","otherDown","thumb-down"]],["최종 업데이트: 2025-09-04(UTC)"],[],[],null,["# Interfaces for Vertex AI Pipelines\n\n| To learn more,\n| run the \"Learn how to build Python function-based Kubeflow pipeline\n| components\" notebook in one of the following\n| environments:\n|\n| [Open in Colab](https://colab.research.google.com/github/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/official/pipelines/lightweight_functions_component_io_kfp.ipynb)\n|\n|\n| \\|\n|\n| [Open in Colab Enterprise](https://console.cloud.google.com/vertex-ai/colab/import/https%3A%2F%2Fraw.githubusercontent.com%2FGoogleCloudPlatform%2Fvertex-ai-samples%2Fmain%2Fnotebooks%2Fofficial%2Fpipelines%2Flightweight_functions_component_io_kfp.ipynb)\n|\n|\n| \\|\n|\n| [Open\n| in Vertex AI Workbench](https://console.cloud.google.com/vertex-ai/workbench/deploy-notebook?download_url=https%3A%2F%2Fraw.githubusercontent.com%2FGoogleCloudPlatform%2Fvertex-ai-samples%2Fmain%2Fnotebooks%2Fofficial%2Fpipelines%2Flightweight_functions_component_io_kfp.ipynb)\n|\n|\n| \\|\n|\n| [View on GitHub](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/official/pipelines/lightweight_functions_component_io_kfp.ipynb)\n\nThis page lists the interfaces that you can use to define and run ML pipelines\non Vertex AI Pipelines.\n\nInterfaces to define a pipeline\n-------------------------------\n\nVertex AI Pipelines supports ML pipelines defined using the\nKubeflow Pipelines (KFP) SDK or the TensorFlow Extended (TFX) SDK.\n\n### Kubeflow Pipelines (KFP) SDK\n\n\nUse KFP for all use\ncases where you don't need to use TensorFlow Extended to process huge amounts of\nstructured or text data. Vertex AI Pipelines supports KFP SDK v2.0 or later.\n\nWhen you use the KFP SDK, you can define your ML workflow by building custom\ncomponents and also by reusing prebuilt components, such as the\nGoogle Cloud Pipeline Components. Google Cloud Pipeline Components let you easily use Vertex AI\nservices like AutoML in your ML pipeline. Vertex AI Pipelines\nsupports Google Cloud Pipeline Components SDK v2 or later. For more information about\nGoogle Cloud Pipeline Components, see\n[Introduction to Google Cloud Pipeline Components](/vertex-ai/docs/pipelines/components-introduction).\n\nTo learn how to build a pipeline using the Kubeflow Pipelines, see\n[Build a pipeline](/vertex-ai/docs/pipelines/build-pipeline). To learn more about\nKubeflow Pipelines, see the [Kubeflow Pipelines documentation](https://www.kubeflow.org/docs/components/pipelines/).\n\n### TensorFlow Extended (TFX) SDK\n\n\nUse TFX if you use TensorFlow Extended in your ML workflow to process\nterabytes of structured or text data. Vertex AI Pipelines supports\nTFX SDK v0.30.0 or later.\n\nTo learn how to build ML pipelines using TFX, see the\n[Getting started tutorials](https://www.tensorflow.org/tfx/tutorials)\nsection on the [TensorFlow Extended in Production tutorials](https://www.tensorflow.org/tfx/tutorials#tfx-on-google-cloud).\n\nInterfaces to run a pipeline\n----------------------------\n\nAfter you define your ML pipeline, you can create an ML pipeline run using any\nof the following interfaces:\n\n- REST API\n\n- SDK clients\n\n- Google Cloud console\n\n| **Note:** Vertex AI Pipelines doesn't support the gcloud CLI interface.\n\nFor more information about the interfaces you can use to interact with Vertex AI, see [Interfaces for Vertex AI](/vertex-ai/docs/start/introduction-interfaces).\n\n### REST API\n\nTo create a pipeline run using REST, use the `Pipelines` service API. This API uses the [`projects.locations.pipelineJobs`](/vertex-ai/docs/reference/rest/v1/projects.locations.pipelineJobs) REST resource.\n| **Note:** Breaking changes to the `Pipelines` service API are communicated as preview launches. You can test the changes announced in preview, see the the API documentation for [`projects.locations.pipelineJobs` (v1beta1)](/vertex-ai/docs/reference/rest/v1beta1/projects.locations.pipelineJobs). For more information about the preview launch stage, see the [launch stage descriptions](/products#product-launch-stages).\n\n### SDK Clients\n\nVertex AI Pipelines lets you create pipeline runs using the Vertex AI SDK for Python or client libraries.\n\n#### Vertex AI SDK for Python\n\nThe Vertex AI SDK for Python (`aiplatform`) is the recommended SDK for programmatically working with the `Pipelines` service API. For more information about this SDK, see the [API documentation for `google.cloud.aiplatform.PipelineJob`](/python/docs/reference/aiplatform/latest/google.cloud.aiplatform/google.cloud.aiplatform.PipelineJob).\n\n#### Client libraries\n\nClient libraries are programmatically Generated API Clients (GAPIC) SDKs. Vertex AI Pipelines supports the following client libraries:\n\n- Python (`aiplatform` `v1` and `v1beta1`)\n\n- Java\n\n- Node.js\n\n- [Go](https://cloud.google.com/go/docs/reference/cloud.google.com/go/aiplatform/latest/apiv1#cloud_google_com_go_aiplatform_apiv1_PipelineClient)\n\nFor more information, see [Install the Vertex AI client libraries](/vertex-ai/docs/start/client-libraries).\n\n### Google Cloud console (GUI)\n\nGoogle Cloud console is the recommended way for reviewing and monitoring your pipeline runs. You can also perform other tasks using the Google Cloud console, such as creating, deleting and cloning pipeline runs, accessing the Template Gallery, and retrieving the billing label for a pipeline run.\n\n[Go to Pipelines in Google Cloud console](https://console.cloud.google.com/vertex-ai/pipelines)\n\nWhat's next\n-----------\n\n- Get started by [learning how to define a pipeline using the Kubeflow Pipelines SDK](/vertex-ai/docs/pipelines/build-pipeline).\n\n- [Learn how to run a pipeline](/vertex-ai/docs/pipelines/run-pipeline).\n\n- Learn about [best practices for implementing custom-trained ML models on Vertex AI](https://cloud.google.com/architecture/ml-on-gcp-best-practices#machine-learning-workflow-orchestration)."]]
대규모 정형 데이터 또는 텍스트 데이터를 처리하기 위해 TensorFlow Extended를 사용할 필요가 없는 모든 사용 사례에 KFP를 사용하세요.
ML 워크플로에서 TensorFlow Extended를 사용하여 테라바이트 규모의 정형 데이터 또는 텍스트 데이터를 처리하는 경우 TFX를 사용하세요. Vertex AI Pipelines는 TFX SDK v0.30.0 이상을 지원합니다.