BatchPredictionJob 리소스를 사용하면 비동기 예측 요청을 실행할 수 있습니다. model 리소스에서 바로 일괄 예측을 요청합니다. 모델을 endpoint에 배포할 필요가 없습니다. 일괄 및 온라인 예측을 지원하는 데이터 유형의 경우 일괄 예측을 사용할 수 있습니다.
이것은 즉각적인 응답이 필요하지 않고 단일 요청을 사용하여 누적된 데이터를 처리하려는 경우에 유용합니다.
일괄 예측을 수행하려면 Vertex AI가 예측 결과를 저장하는 입력 소스와 출력 위치를 지정합니다. 입력과 출력은 작업 중인 model 유형에 따라 다릅니다. 예를 들어 AutoML 이미지 모델 유형의 일괄 예측에는 입력 JSON Lines 파일과 출력을 저장할 Cloud Storage 버킷 이름이 필요합니다.
일괄 예측에 대한 자세한 내용은 일괄 예측 가져오기를 참조하세요.
[[["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: 2025-06-23(UTC)"],[],[],null,["# Batch prediction components\n\n| To learn more,\n| run the \"Learn how to use prebuilt Pipeline Components to train a custom model\" 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/custom_model_training_and_batch_prediction.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%2Fcustom_model_training_and_batch_prediction.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%2Fcustom_model_training_and_batch_prediction.ipynb)\n|\n|\n| \\|\n|\n| [View on GitHub](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/official/pipelines/custom_model_training_and_batch_prediction.ipynb)\n\nThe `BatchPredictionJob` resource lets you run an asynchronous\nprediction request. Request batch predictions directly from the `model`\nresource. You don't need to deploy the model to an `endpoint`. For data types\nthat support both batch and online predictions you can use batch predictions.\nThis is useful when you don't require an immediate response and want to process\naccumulated data by using a single request.\n\nTo make a batch prediction, specify an input source and an output location\nfor Vertex AI to store predictions results. The inputs and outputs\ndepend on the `model` type that you're working with. For example, batch\npredictions for the AutoML image model type require an input\n[JSON Lines](https://jsonlines.org/)\nfile and the name of a Cloud Storage bucket to store the output.\nFor more information about batch prediction, see\n[Get batch predictions](/vertex-ai/docs/predictions/batch-predictions).\n\nYou can use the [`ModelBatchPredictOp`](https://google-cloud-pipeline-components.readthedocs.io/en/google-cloud-pipeline-components-2.19.0/api/v1/batch_predict_job.html#v1.batch_predict_job.ModelBatchPredictOp) component to access this resource through Vertex AI Pipelines.\n\nAPI reference\n-------------\n\n- For component reference, see the [Google Cloud SDK reference for Batch prediction components](https://google-cloud-pipeline-components.readthedocs.io/en/google-cloud-pipeline-components-2.19.0/api/v1/batch_predict_job.html).\n- For Vertex AI API reference, see the [`BatchPredictionJob` resource](/vertex-ai/docs/reference/rest/v1/projects.locations.batchPredictionJobs) page.\n\nTutorials\n---------\n\n- [Custom training with prebuilt Google Cloud Pipeline Components](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/official/pipelines/custom_model_training_and_batch_prediction.ipynb)\n\n### Version history and release notes\n\nTo learn more about the version history and changes to the Google Cloud Pipeline Components SDK, see the [Google Cloud Pipeline Components SDK Release Notes](https://google-cloud-pipeline-components.readthedocs.io/en/google-cloud-pipeline-components-2.19.0/release.html).\n\n### Technical support contacts\n\nIf you have any questions, reach out to\n[kubeflow-pipelines-components@google.com](mailto: kubeflow-pipelines-components@google.com)."]]
