“项目”是指您的 Google Cloud 项目。它是您部署的模型和作业的逻辑容器。您用于开发 AI Platform Prediction 解决方案的每个项目都必须已启用 AI Platform Prediction。您的 Google 账号可以访问多个 Google Cloud项目
。
型号
在 ML 中，“模型”代表着您尝试解决的问题的解决方案。换句话说，它是根据数据预测值的配方。在 AI Platform Prediction 中，模型是这种解决方案各个版本的逻辑容器。例如，假设您要解决的问题是根据一组有关先前销售的数据预测房屋的销售价格。在 AI Platform Prediction 中创建一个名为 housing_prices 的模型，并尝试使用多种机器学习技术来解决问题。在每个阶段，您都可以部署该模型的版本。每个版本都可能与其他版本截然不同，但您可以将这些版本组织在同一模型下（只要这种做法适合您的工作流）。
模型版本（或者简称为版本）是存储在 AI Platform Prediction 模型服务中的机器学习解决方案的实例。您通过将序列化且经过训练的模型（作为已保存的模型）传递给服务来创建版本。当您创建版本时，您还可以提供自定义代码（Beta 版）来处理预测。
作业
您可以通过发起请求和作业来与 AI Platform Prediction 的服务进行交互。请求属于常规 Web API 请求，会尽快返回响应。作业是长时间运行的异步处理操作。AI Platform Prediction 提供训练作业和批量预测作业。您提交启动作业的请求并获得快速响应，以验证作业状态。随后，您可以定期请求状态以跟踪作业的进度。
[[["易于理解","easyToUnderstand","thumb-up"],["解决了我的问题","solvedMyProblem","thumb-up"],["其他","otherUp","thumb-up"]],[["很难理解","hardToUnderstand","thumb-down"],["信息或示例代码不正确","incorrectInformationOrSampleCode","thumb-down"],["没有我需要的信息/示例","missingTheInformationSamplesINeed","thumb-down"],["翻译问题","translationIssue","thumb-down"],["其他","otherDown","thumb-down"]],["最后更新时间 (UTC)：2025-04-11。"],[[["\u003cp\u003eA project is your Google Cloud project, serving as the logical container for deployed models and jobs, and requiring AI Platform Prediction to be enabled for AI Platform Prediction solution development.\u003c/p\u003e\n"],["\u003cp\u003eIn AI Platform Prediction, a model is a logical container for different versions of a machine learning solution, each representing the recipe for predicting a value from data, such as predicting the sale price of houses.\u003c/p\u003e\n"],["\u003cp\u003eA model version, or simply version, is a specific instance of a machine learning solution stored in the AI Platform Prediction service, created by passing a serialized trained model (saved model) to the service.\u003c/p\u003e\n"],["\u003cp\u003eJobs are long-running, asynchronous operations within AI Platform Prediction, including training and batch prediction jobs, initiated by a request and tracked through status updates.\u003c/p\u003e\n"],["\u003cp\u003eDeploying a model version involves creating a version resource by specifying an exported model (saved model directory) and a model resource, allowing AI Platform Prediction to host and provide predictions from it.\u003c/p\u003e\n"]]],[],null,[]]
