[[["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):2025-04-14。"],[[["AlloyDB can be used as a vector database by utilizing the `vector` extension, which includes `pgvector` functions and operators for storing embeddings."],["The `vector` extension, version `0.5.0.google-1` or later, is required and has Google-specific optimizations for AlloyDB to efficiently manage vector values."],["To store embeddings, add a `vector[]` column to your table with `ALTER TABLE`, specifying the number of dimensions supported by your model, and then populate it with generated embeddings."],["If using the Langchain framework with a dataset of 100k embeddings or more, consider using the `AlloyDBVectorStore` vector class from the alloydb langChain library."],["Once embeddings are stored, you can use either the `vector` extension or the `alloydb_scann` extension to create indexes, which will improve query performance."]]],[]]