[[["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):2025-04-02。"],[[["Continuous evaluation samples input and output from deployed machine learning models to provide ongoing feedback on model performance."],["Evaluation jobs are created for deployed model versions, which then save prediction data to a BigQuery table and run periodically to generate evaluation metrics."],["Ground truth labels, which serve as an answer key, can either be provided by the Data Labeling Service or by the user directly, for comparing against model predictions."],["Evaluation job runs occur daily by default and create datasets for evaluation, with the timing of metric calculations depending on the ground truth labeling method chosen."],["Continuous evaluation requires the use of AI Platform Prediction, BigQuery and Cloud Storage, and if human review is chosen as a ground truth labeling method, then Data Labeling Service pricing will apply."]]],[]]