{"id":42107,"date":"2023-07-14T06:46:25","date_gmt":"2023-08-19T07:49:36","guid":{"rendered":"https:\/\/www.silicloud.com\/zh\/blog\/%e4%bd%bf%e7%94%a8bedrock%e5%92%8caurora-serverless-postgresql%e7%9a%84pgvector%e6%9d%a5%e8%bf%9b%e8%a1%8crag%ef%bc%88retrieve-aggregate-generate%ef%bc%89-%e5%9c%a8amazon-titan-embeddings%e4%b8%ad\/"},"modified":"2024-04-30T13:30:23","modified_gmt":"2024-04-30T05:30:23","slug":"%e4%bd%bf%e7%94%a8bedrock%e5%92%8caurora-serverless-postgresql%e7%9a%84pgvector%e6%9d%a5%e8%bf%9b%e8%a1%8crag%ef%bc%88retrieve-aggregate-generate%ef%bc%89-%e5%9c%a8amazon-titan-embeddings%e4%b8%ad","status":"publish","type":"post","link":"https:\/\/www.silicloud.com\/zh\/blog\/%e4%bd%bf%e7%94%a8bedrock%e5%92%8caurora-serverless-postgresql%e7%9a%84pgvector%e6%9d%a5%e8%bf%9b%e8%a1%8crag%ef%bc%88retrieve-aggregate-generate%ef%bc%89-%e5%9c%a8amazon-titan-embeddings%e4%b8%ad\/","title":{"rendered":"\u4f7f\u7528Bedrock\u548cAurora Serverless PostgreSQL\u7684pgvector\u6765\u8fdb\u884cRAG\uff08Retrieve, Aggregate, Generate\uff09- \u5728Amazon Titan Embeddings\u4e2d\u5b66\u4e60\u5411\u91cf"},"content":{"rendered":"

\u6211\u4eec\u5c06\u4f7f\u7528Aurora Serverless PostgreSQL\u7684pgvector\u4f5c\u4e3aVector DB\u6765\u6784\u5efaRAG\u3002<\/p>\n

\u6982\u5ff5\u56fe<\/h1>\n
\"image.png\"<\/div>\n

\u5efa\u7b51\u84dd\u56fe<\/h1>\n
\"image.png\"<\/div>\n

\u65e0\u9700\u914d\u7f6e\u7684Aurora Serverless PostgreSQL<\/h1>\n
\"image.png\"<\/div>\n
\"image.png\"<\/div>\n
\"image.png\"<\/div>\n
\"image.png\"<\/div>\n
\"image.png\"<\/div>\n
\"image.png\"<\/div>\n
\"image.png\"<\/div>\n

Cloud 9\u7684\u6784\u6210<\/h2>\n

\u521b\u5efa\u4e00\u4e2a\u540d\u4e3aCloud9\u7684\u5b9e\u4f8b\u3002\u9ed8\u8ba4\u8bbe\u7f6e\u5373\u53ef\u3002
\n\u5c06Cloud9\u7684\u5b89\u5168\u7ec4\u6dfb\u52a0\u5230PostgreSQL\u7684\u5b89\u5168\u7ec4\uff08\u9ed8\u8ba4\u5b89\u5168\u7ec4\uff09\u7684\u5165\u7ad9\u89c4\u5219\u4e2d\u3002\u6682\u65f6\u5141\u8bb8\u6240\u6709\u7684\u6d41\u91cf\u3002<\/p>\n

\u5b89\u88c5PostgreSQL\u5ba2\u6237\u7aef\u3002<\/h3>\n
sudo amazon-linux-extras install postgresql14\r\n<\/code><\/pre>\n
\u542f\u7528pgvector<\/h2>\n
\u6211\u5c06\u4f7f\u7528psql\u8fde\u63a5\u5230\u7f16\u5199\u5668\u7aef\u70b9\u3002<\/p>\n
psql -h vector-test.cluster-xxxxxxxxxxxx.us-east-1.rds.amazonaws.com -U postgres -p 5432\r\n<\/code><\/pre>\n
\u8bf7\u8f93\u5165\u5bc6\u7801\u5e76\u6309Enter\u952e\u3002<\/p>\n
Password for user postgres: \r\npsql (14.8, server 15.4)\r\nWARNING: psql major version 14, server major version 15.\r\n         Some psql features might not work.\r\nSSL connection (protocol: TLSv1.2, cipher: AES128-SHA256, bits: 128, compression: off)\r\nType \"help\" for help.\r\n\r\npostgres=> \r\n<\/code><\/pre>\n
\u53ea\u8981\u662f\u8fd9\u6b21\u7684\u7528\u9014\uff0c\u4e0d\u5fc5\u62c5\u5fc3\u8b66\u544a\uff0c\u6ca1\u95ee\u9898\u7684\u3002
\n\u7136\u540e\u6309\u7167\u4ee5\u4e0b\u65b9\u5f0f\u8f93\u5165\uff0c\u5f00\u542f\u4e86pgvector\u3002<\/p>\n
postgres=> CREATE EXTENSION vector;\r\nCREATE EXTENSION\r\n<\/code><\/pre>\n
\u7531\u4e8eLangChain\u5df2\u521b\u5efa\u4e86\u8868\u683c\u7b49\u5185\u5bb9\uff0c\u56e0\u6b64\u65e0\u9700\u5728\u6b64\u5904\u8fdb\u884c\u521b\u5efa\u3002\u4e4b\u540e\u53ea\u9700\u8981\u786e\u8ba4\u5e76\u79bb\u5f00\u5373\u53ef\u3002<\/p>\n
postgres=> \\dx\r\n                        List of installed extensions\r\n  Name   | Version |   Schema   |                Description                 \r\n---------+---------+------------+--------------------------------------------\r\n plpgsql | 1.0     | pg_catalog | PL\/pgSQL procedural language\r\n vector  | 0.5.0   | public     | vector data type and ivfflat access method\r\n(2 rows)\r\n\r\npostgres=> exit\r\n<\/code><\/pre>\n
Python\u73af\u5883\u7684\u914d\u7f6e<\/h1>\n
\u5b89\u88c5\u5fc5\u8981\u7684\u5e93\u3002<\/h2>\n
pip install boto3\r\npip install langchain\r\npip install pgvector\r\npip install psycopg2-binary\r\npip install pypdf\r\npip install streamlit\r\n<\/code><\/pre>\n
\u521b\u5efa\u5411\u91cf\u6570\u636e\u5e93<\/h2>\n
\u5c06\u8fd9\u4e2aPDF\u6587\u4ef6\u5b58\u50a8\u5230PostgreSQL\u4e2d\u3002<\/p>\n
 <\/p>\n
\u9996\u5148\uff0c\u5c06\u4ee5\u4e0a\u6587\u4ef6\u4e0a\u4f20\u5230Cloud9\u3002
\n\u5728\u8fd9\u4e2a\u4f8b\u5b50\u4e2d\uff0c\u6211\u4eec\u5c06\u5176\u653e\u7f6e\u5728\/home\/ec2-user\/environment\/bedrock-ug.pdf\u4f4d\u7f6e\u4e0a\u3002<\/p>\n
\u6211\u5c06\u521b\u5efa\u5e76\u8fd0\u884c\u4ee5\u4e0b\u7a0b\u5e8f\u3002
\n\u8bf7\u8bbe\u7f6eCONNECTION_STRING\u7684{password}\u4e3aPostgreSQL\u7684\u4e3b\u5bc6\u7801\u3002\u8bf7\u8bbe\u7f6evector-test.cluster-xxxxxxxxxxxx.us-east-1.rds.amazonaws.com\u4f5c\u4e3a\u8bfb\u53d6\u7aef\u70b9\u3002<\/p>\n
from<\/span> langchain.embeddings<\/span> import<\/span> BedrockEmbeddings<\/span>\r\nfrom<\/span> langchain.vectorstores.pgvector<\/span> import<\/span> PGVector<\/span>\r\nfrom<\/span> langchain.document_loaders<\/span> import<\/span> PyPDFLoader<\/span>\r\n\r\n# \u63a5\u7d9a\u6587\u5b57\u5217\u306e\u5b9a\u7fa9\r\n<\/span>CONNECTION_STRING<\/span> =<\/span> \"<\/span>postgresql+psycopg2:\/\/postgres:{password}@vector-test.cluster-xxxxxxxxxxxx.us-east-1.rds.amazonaws.com:5432<\/span>\"<\/span>\r\n\r\n# Embeddings\u306e\u5b9a\u7fa9\r\n<\/span>embeddings<\/span> =<\/span> BedrockEmbeddings<\/span>(<\/span>model_id<\/span>=<\/span>\"<\/span>amazon.titan-embed-text-v1<\/span>\"<\/span>)<\/span>\r\n\r\n# PostgreSQL\u306e\u5b9a\u7fa9\r\n<\/span>db<\/span> =<\/span> PGVector<\/span>(<\/span>\r\n    connection_string<\/span>=<\/span>CONNECTION_STRING<\/span>,<\/span>\r\n    embedding_function<\/span>=<\/span>embeddings<\/span>\r\n)<\/span>\r\n\r\n# PDF\u30d5\u30a1\u30a4\u30eb\u3092DB\u306b\u8ffd\u52a0\r\n<\/span>loader<\/span> =<\/span> PyPDFLoader<\/span>(<\/span>r<\/span>\"<\/span>\/home\/ec2-user\/environment\/bedrock-ug.pdf<\/span>\"<\/span>)<\/span>\r\npages<\/span> =<\/span> loader<\/span>.<\/span>load_and_split<\/span>()<\/span>\r\ndb<\/span>.<\/span>add_documents<\/span>(<\/span>documents<\/span>=<\/span>pages<\/span>)<\/span>\r\n<\/code><\/pre>\n
\u8fd9\u4e2a\u8fc7\u7a0b\u5e94\u8be5\u4f1a\u5c06\u4e0a\u8ff0\u7684PDF\u6587\u4ef6\u6309\u7167\u4e00\u9875\u4e00\u9875\u5730\u5206\u5272\uff0c\u5e76\u901a\u8fc7Titan Embeddings\u8fdb\u884c\u5d4c\u5165\uff0c\u7136\u540e\u5b58\u50a8\u5230PostgreSQL\u4e2d\u3002<\/p>\n
\u987a\u4fbf\u63d0\u4e00\u4e0b\uff0cPGVector\u7684\u63a5\u53e3\u5982\u4e0b\u3002<\/p>\n
 <\/p>\n
\u786e\u8ba4\u5b58\u50a8\u5185\u5bb9<\/h1>\n
\u4f7f\u7528psql\u8fde\u63a5\u5230leader\u8282\u70b9\u7684\u7ec8\u7aef\u3002<\/p>\n
psql -h vector-test.cluster-xx-xxxxxxxxxxxx.us-east-1.rds.amazonaws.com -U postgres -p 5432\r\n<\/code><\/pre>\n
\u901a\u8fc7\u4f7f\u7528LangChain\u521b\u5efa\u7684\u8868\u683c\u8fdb\u884c\u786e\u8ba4\u3002<\/h2>\n
postgres=> \\d\r\n                  List of relations\r\n Schema |          Name           | Type  |  Owner   \r\n--------+-------------------------+-------+----------\r\n public | langchain_pg_collection | table | postgres\r\n public | langchain_pg_embedding  | table | postgres\r\n(2 rows)\r\n\r\npostgres=> \r\n<\/code><\/pre>\n
\u901a\u8fc7LangChain\u5206\u522b\u521b\u5efa\u4e86langchain_pg_collection\u548clangchain_pg_embedding\u4e24\u4e2a\u8868\u683c\u3002<\/p>\n
\u786e\u8ba4\u6536\u85cf<\/h2>\n
\u786e\u8ba4langchain_pg_collection\u3002<\/p>\n
postgres=> \\d langchain_pg_collection;\r\n             Table \"public.langchain_pg_collection\"\r\n  Column   |       Type        | Collation | Nullable | Default \r\n-----------+-------------------+-----------+----------+---------\r\n name      | character varying |           |          | \r\n cmetadata | json              |           |          | \r\n uuid      | uuid              |           | not null | \r\nIndexes:\r\n    \"langchain_pg_collection_pkey\" PRIMARY KEY, btree (uuid)\r\nReferenced by:\r\n    TABLE \"langchain_pg_embedding\" CONSTRAINT \"langchain_pg_embedding_collection_id_fkey\" FOREIGN KEY (collection_id) REFERENCES langchain_pg_collection(uuid) ON DELETE CASCADE\r\n\r\npostgres=> select * from langchain_pg_collection;\r\n   name    | cmetadata |                 uuid                 \r\n-----------+-----------+--------------------------------------\r\n langchain | null      | 79292fae-56a5-4934-ad94-6e1f2b7a9185\r\n(1 row)\r\n<\/code><\/pre>\n
\u5df2\u521b\u5efa\u4e86\u4e00\u4e2a\u8bb0\u5f55\u3002
\nname\u7684langchain\u5c5e\u6027\u662fLangChain\u7684\u9ed8\u8ba4\u503c\u3002\u56e0\u4e3a\u5728\u6b64\u6b21\u7a0b\u5e8f\u521b\u5efa\u65f6\u6ca1\u6709\u6307\u5b9a\u503c\uff0c\u6240\u4ee5\u53d8\u6210\u4e86\u9ed8\u8ba4\u503c\u3002\u4f46\u4e5f\u53ef\u4ee5\u901a\u8fc7\u53c2\u6570\u6307\u5b9a\u6765\u521b\u5efa\u3002\u540c\u65f6\uff0c\u8be5\u8bb0\u5f55\u8fd8\u88ab\u5206\u914d\u4e86\u4e00\u4e2a\u4f5c\u4e3auuid\u7684ID\uff0c\u88ab\u79f0\u4e3a\u201cCollection\u201d\u3002<\/p>\n
\u786e\u8ba4\u5d4c\u5165\u8868<\/h2>\n
\u6211\u8981\u786e\u8ba4\u4e00\u4e0blangchain_pg_embedding\u3002<\/p>\n
postgres=> \\d langchain_pg_embedding;\r\n               Table \"public.langchain_pg_embedding\"\r\n    Column     |       Type        | Collation | Nullable | Default \r\n---------------+-------------------+-----------+----------+---------\r\n collection_id | uuid              |           |          | \r\n embedding     | vector            |           |          | \r\n document      | character varying |           |          | \r\n cmetadata     | json              |           |          | \r\n custom_id     | character varying |           |          | \r\n uuid          | uuid              |           | not null | \r\nIndexes:\r\n    \"langchain_pg_embedding_pkey\" PRIMARY KEY, btree (uuid)\r\nForeign-key constraints:\r\n    \"langchain_pg_embedding_collection_id_fkey\" FOREIGN KEY (collection_id) REFERENCES langchain_pg_collection(uuid) ON DELETE CASCADE\r\n\r\npostgres=> \r\n<\/code><\/pre>\n
\u6211\u4f1a\u786e\u8ba4\u6570\u91cf\u3002<\/p>\n
postgres=> select count(*) from langchain_pg_embedding;\r\n count \r\n-------\r\n   156\r\n(1 row)\r\n\r\npostgres=> \r\n<\/code><\/pre>\n
\u53ef\u4ee5\u770b\u5230\u5df2\u521b\u5efa\u4e0ePDF\u6587\u4ef6\u9875\u6570\u76f8\u5bf9\u5e94\u7684\u8bb0\u5f55\u3002
\n\u7a0d\u5fae\u67e5\u770b\u4e00\u4e0b\u3002\u7531\u4e8e\u5d4c\u5165\u662f\u5b58\u50a8\u5411\u91cf\u503c\u7684\u5de8\u5927\u5217\uff0c\u6240\u4ee5\u9996\u5148\u8fdb\u884c\u6392\u9664\u641c\u7d22\u3002<\/p>\n
postgres=> select collection_id,document,custom_id,uuid from langchain_pg_embedding;\r\n            collection_id             |                                                                         document                                               \r\n                           |              custom_id               |                 uuid                 \r\n--------------------------------------+--------------------------------------------------------------------------------------------------------------------------------\r\n---------------------------+--------------------------------------+--------------------------------------\r\n 79292fae-56a5-4934-ad94-6e1f2b7a9185 | Amazon Bedrock                                                                                                                 \r\n                          +| d5f0a7e4-7bc8-11ee-8460-06a674c2adb7 | 28d92844-45e8-4880-a6f8-79379a981e62\r\n                                      | \u30e6\u30fc\u30b6\u30fc\u30ac\u30a4\u30c9                                                                                                                 \r\n                           |                                      | \r\n 79292fae-56a5-4934-ad94-6e1f2b7a9185 | Amazon Bedrock \u30e6\u30fc\u30b6\u30fc\u30ac\u30a4\u30c9                                                                                                  \r\n                          +| d5f0a992-7bc8-11ee-8460-06a674c2adb7 | 90199c78-cd20-4143-adc7-6d11e474a972\r\n                                      | Amazon Bedrock: \u30e6\u30fc\u30b6\u30fc\u30ac\u30a4\u30c9                                                                                                 \r\n                          +|                                      | \r\n                                      | Copyright \u00a9 2023 Amazon Web Services, Inc. and\/or its affiliates. All rights reserved.                                         \r\n                          +|                                      | \r\n                                      | Amazon \u306e\u5546\u6a19\u304a\u3088\u3073\u30c8\u30ec\u30fc\u30c9\u30c9\u30ec\u30b9\u306f\u3001\u9867\u5ba2\u306b\u6df7\u4e71\u3092\u62db\u304f\u53ef\u80fd\u6027\u304c\u3042\u308b\u614b\u69d8\u3001\u307e\u305f\u306f Amazon \u306e\u4fe1\u7528\u3092\u50b7\u3064\u3051                            \r\n                          +|                                      | \r\n                                      | \u305f\u308a\u3001\u5931\u308f\u305b\u305f\u308a\u3059\u308b\u614b\u69d8\u306b\u304a\u3044\u3066\u3001Amazon \u306e\u3082\u306e\u3067\u306f\u306a\u3044\u88fd\u54c1\u307e\u305f\u306f\u30b5\u30fc\u30d3\u30b9\u306b\u95a2\u9023\u3057\u3066\u4f7f\u7528\u3057\u3066\u306f\u306a\u308a\u307e\u305b                          \r\n                          +|                                      | \r\n                                      | \u3093\u3002Amazon \u304c\u6240\u6709\u3057\u306a\u3044\u5546\u6a19\u306f\u3059\u3079\u3066\u305d\u308c\u305e\u308c\u306e\u6240\u6709\u8005\u306b\u6240\u5c5e\u3057\u307e\u3059\u3002\u6240\u6709\u8005\u306f\u5fc5\u305a\u3057\u3082 Amazon \u3068\u306e\u63d0\u643a\u3084\u95a2\u9023                        \r\n                          +|                                      | \r\n                                      | \u304c\u3042\u308b\u308f\u3051\u3067\u306f\u3042\u308a\u307e\u305b\u3093\u3002\u307e\u305f\u3001Amazon \u306e\u652f\u63f4\u3092\u53d7\u3051\u3066\u3044\u308b\u3068\u306f\u9650\u308a\u307e\u305b\u3093\u3002                                                      \r\n                           |                                      | \r\n 79292fae-56a5-4934-ad94-6e1f2b7a9185 | Amazon Bedrock \u30e6\u30fc\u30b6\u30fc\u30ac\u30a4\u30c9                                                                                                  \r\n                          +| d5f0aa50-7bc8-11ee-8460-06a674c2adb7 | ee5cde36-2e92-4b2a-9f7d-415032315a6b\r\n                                      | Table of Contents                                                                                                              \r\n                          +|                                      | \r\n                                      | \u30a2\u30de\u30be\u30f3\u30fb\u30d9\u30c3\u30c9\u30ed\u30c3\u30af\u3068\u306f\uff1f ..................................................................................................\r\n...........1              +|                                      | \r\n                                      | Bedrock \u30e2\u30c7\u30eb\u306b\u30a2\u30af\u30bb\u30b9\u3057\u3066\u304f\u3060\u3055\u3044\u3002 .....................................................................................1  \r\n                          +|                                      | \r\n<\/code><\/pre>\n
\u5c3d\u7ba1\u6587\u5b57\u6570\u8f83\u591a\uff0c\u53ef\u80fd\u6709\u4e9b\u96be\u4ee5\u7406\u89e3\uff0c\u4f46collection_id\u4e2d\u5b58\u50a8\u4e86\u96c6\u5408\u7684uuid\u3002
\ndocument\u4e2d\u5b58\u50a8\u4e86\u6bcf\u4e2a\u9875\u9762\u7684\u6587\u672c\u3002
\nuuid\u662f\u4e3a\u6bcf\u6761\u8bb0\u5f55\u5206\u914d\u7684\u4e0d\u540cID\u3002<\/p>\n
\u67e5\u770b\u5d4c\u5165\u5185\u5bb9\u4e2d\u5b58\u50a8\u7684\u4fe1\u606f\u3002<\/p>\n
postgres=> select embedding from langchain_pg_embedding;\r\n\r\n[0.18457031,0.38671875,0.484375,-0.022338867,0.7578125,0.63671875,-0.18652344,-0.00037384033,0.012207031,-0.3515625,-0.095703125,0.69140625,0.4765625,0.140625,-0.3710\r\n9375,-0.083984375,-0.47460938,0.29492188,-0.30078125,-0.029541016,-0.76171875,0.98828125,0.053466797,0.9453125,-0.42773438,0.625,0.08691406,-0.37890625,-0.609375,-0.31\r\n054688,-0.75,0.65234375,-0.2578125,-0.56640625,-0.20019531,-0.12890625,0.19726562,-1.1015625,-0.65234375,0.69921875,-0.0859375,-0.5546875,-0.123046875,0.21386719,0.279\r\n29688,-0.36132812,0.07128906,0.29101562,1.171875,0.67578125,-0.03149414,0.0026855469,0.21484375,0.30273438,0.16796875,0.26367188,0.50390625,-0.47265625,0.65625,0.26757\r\n812,-0.024169922,0.35546875,-0.13085938,0.25390625,0.4296875,0.0859375,0.80078125,-0.30859375,-0.52734375,0.92578125,-0.671875,0.94140625,0.27148438,-0.2265625,-0.0986\r\n3281,0.48632812,-0.41015625,0.47070312,0.828125,0.056152344,-0.21972656,0.5703125,0.20214844,0.29101562,-0.39257812,-0.51953125,-0.45898438,0.29296875,0.00013828278,\r\n<\/code><\/pre>\n
\u56e0\u4e3a\u6ca1\u6709\u610f\u4e49\uff0c\u6240\u4ee5\u6211\u4f1a\u5728\u4e2d\u9014\u505c\u4e0b\u6765\uff0c\u4f46\u662f\u53ef\u4ee5\u770b\u51fa\u5d4c\u5165\u4e2d\u5b58\u50a8\u4e86\u5411\u91cf\u503c\u3002
\n\u6362\u53e5\u8bdd\u8bf4\uff0c\u5f53\u8fdb\u884c\u5411\u91cf\u641c\u7d22\u65f6\uff0c\u53ef\u4ee5\u4ece\u5d4c\u5165\u4e2d\u83b7\u53d6\u76f8\u4f3c\u5ea6\u8f83\u9ad8\u7684\u8bb0\u5f55\uff0c\u5e76\u83b7\u53d6\u5230\u5b9e\u9645\u7684\u6587\u6863\u5185\u5bb9\u3002<\/p>\n
RAG\u7684\u5b9e\u65bd<\/h1>\n
\u521b\u5efa\u5e76\u6267\u884c\u4ee5\u4e0b\u7a0b\u5e8f\u3002
\n\u4e0e\u4e4b\u524d\u76f8\u540c\uff0c\u6211\u4eec\u5c06\u4f7f\u7528VectorDB\u548cTitan Embeddings\u4f5c\u4e3a\u641c\u7d22\u5185\u5bb9\u7684\u5d4c\u5165\uff0c\u5e76\u4f7f\u7528Claude2\u8fdb\u884c\u6700\u540e\u7684\u53e5\u5b50\u6574\u7406\u3002
\n\u4e3a\u4e86\u65b9\u4fbf\u6267\u884c\uff0c\u6211\u4eec\u4f7f\u7528\u4e86Streamlit\u4f5c\u4e3a\u5916\u5c42\u6846\u67b6\u3002<\/p>\n
\u9664\u7a0b\u5e8f\u7684\u5b9a\u4e49\u90e8\u5206\u5916\uff0c\u5176\u4ed6\u90e8\u5206\u4e0e\u4e0a\u4e00\u7248\u7684ChromaDB\u7a0b\u5e8f\u6ca1\u6709\u53d8\u5316\u3002<\/p>\n
\u8bf7\u8bbe\u7f6e CONNECTION_STRING \u4e2d\u7684 {password} \u4e3a PostgreSQL \u7684\u4e3b\u5bc6\u7801\u3002\u8bf7\u8bbe\u7f6e vector-test.cluster-xx-xxxxxxxxxxxx.us-east-1.rds.amazonaws.com \u4f5c\u4e3a\u8bfb\u53d6\u7aef\u70b9\u3002<\/p>\n
from<\/span> langchain.embeddings<\/span> import<\/span> BedrockEmbeddings<\/span>\r\nfrom<\/span> langchain.vectorstores.pgvector<\/span> import<\/span> PGVector<\/span>\r\nfrom<\/span> langchain.llms<\/span> import<\/span> Bedrock<\/span>\r\nfrom<\/span> langchain.chains<\/span> import<\/span> RetrievalQA<\/span>\r\nfrom<\/span> langchain.prompts<\/span> import<\/span> PromptTemplate<\/span>\r\nimport<\/span> streamlit<\/span> as<\/span> st<\/span>\r\n\r\n# \u63a5\u7d9a\u6587\u5b57\u5217\u306e\u5b9a\u7fa9\r\n<\/span>CONNECTION_STRING<\/span> =<\/span> \"<\/span>postgresql+psycopg2:\/\/postgres:{password}@vector-test.cluster-xx-xxxxxxxxxxxx.us-east-1.rds.amazonaws.com:5432<\/span>\"<\/span>\r\n\r\n# Embeddings\u306e\u5b9a\u7fa9\r\n<\/span>embeddings<\/span> =<\/span> BedrockEmbeddings<\/span>(<\/span>model_id<\/span>=<\/span>\"<\/span>amazon.titan-embed-text-v1<\/span>\"<\/span>)<\/span>\r\n\r\n# PostgreSQL\u306e\u5b9a\u7fa9\r\n<\/span>db<\/span> =<\/span> PGVector<\/span>(<\/span>\r\n    connection_string<\/span>=<\/span>CONNECTION_STRING<\/span>,<\/span>\r\n    embedding_function<\/span>=<\/span>embeddings<\/span>\r\n)<\/span>\r\n\r\n# LLM\u306e\u5b9a\u7fa9\r\n<\/span>llm<\/span> =<\/span> Bedrock<\/span>(<\/span>\r\n    model_id<\/span>=<\/span>\"<\/span>anthropic.claude-v2<\/span>\"<\/span>,<\/span>\r\n    model_kwargs<\/span>=<\/span>{<\/span>\"<\/span>max_tokens_to_sample<\/span>\"<\/span>:<\/span> 1000<\/span>},<\/span>\r\n)<\/span>\r\n\r\n# prompt\u306e\u5b9a\u7fa9\r\n<\/span>prompt_template<\/span> =<\/span> \"\"\"<\/span>\r\n  <documents>{context}<\/documents>\r\n  <\/span>\\n\\n<\/span>Human: \u4e0a\u8a18\u306e\u5185\u5bb9\u3092\u53c2\u8003\u6587\u66f8\u3068\u3057\u3066\u3001\u8cea\u554f\u306e\u5185\u5bb9\u306b\u5bfe\u3057\u3066\u8a73\u3057\u304f\u8aac\u660e\u3057\u3066\u304f\u3060\u3055\u3044\u3002\u8a00\u8a9e\u306e\u6307\u5b9a\u304c\u7121\u3044\u5834\u5408\u306f\u65e5\u672c\u8a9e\u3067\u7b54\u3048\u3066\u304f\u3060\u3055\u3044\u3002\r\n    \u3082\u3057\u8cea\u554f\u306e\u5185\u5bb9\u304c\u53c2\u8003\u6587\u66f8\u306b\u7121\u304b\u3063\u305f\u5834\u5408\u306f\u300c\u6587\u66f8\u306b\u3042\u308a\u307e\u305b\u3093\u300d\u3068\u7b54\u3048\u3066\u304f\u3060\u3055\u3044\u3002\u56de\u7b54\u5185\u5bb9\u306b\u306f\u8cea\u554f\u81ea\u4f53\u3084\u30bf\u30b0\u306f\u542b\u3081\u306a\u3044\u3067\u304f\u3060\u3055\u3044\u3002\r\n    Take a deep breath and work on this problem step-by-step.\r\n  <question>{question}<\/question>\r\n  <\/span>\\n\\n<\/span>Assistant:<\/span>\"\"\"<\/span>\r\nPROMPT<\/span> =<\/span> PromptTemplate<\/span>(<\/span>template<\/span>=<\/span>prompt_template<\/span>,<\/span> input_variables<\/span>=<\/span>[<\/span>\"<\/span>context<\/span>\"<\/span>,<\/span> \"<\/span>question<\/span>\"<\/span>])<\/span>\r\nchain_type_kwargs<\/span> =<\/span> {<\/span>\"<\/span>prompt<\/span>\"<\/span>:<\/span> PROMPT<\/span>}<\/span>\r\n\r\n# Chain\u306e\u5b9a\u7fa9\u3002\u691c\u7d22\u7d50\u679c\u306e\u4e0a\u4f4d15\u4ef6\u3092\u4f7f\u7528\r\n<\/span>qa<\/span> =<\/span> RetrievalQA<\/span>.<\/span>from_chain_type<\/span>(<\/span>llm<\/span>,<\/span>retriever<\/span>=<\/span>db<\/span>.<\/span>as_retriever<\/span>(<\/span>search_kwargs<\/span>=<\/span>{<\/span>\"<\/span>k<\/span>\"<\/span>:<\/span> 15<\/span>}),<\/span>chain_type_kwargs<\/span>=<\/span>chain_type_kwargs<\/span>)<\/span>\r\n\r\n# Streamlit\r\n<\/span>st<\/span>.<\/span>title<\/span>(<\/span>\"<\/span>PostgreSQL PGVectoRAG<\/span>\"<\/span>)<\/span>\r\ninput_text<\/span> =<\/span> st<\/span>.<\/span>text_input<\/span>(<\/span>\"<\/span>\u5165\u529b\u3055\u308c\u305f\u6587\u5b57\u5217\u3067VectorDB\u3092\u691c\u7d22\u3057\u56de\u7b54\u3057\u307e\u3059<\/span>\"<\/span>)<\/span>\r\nsend_button<\/span> =<\/span> st<\/span>.<\/span>button<\/span>(<\/span>\"<\/span>\u9001\u4fe1<\/span>\"<\/span>)<\/span>\r\n\r\nif<\/span> send_button<\/span>:<\/span>\r\n\r\n    # \u5b9f\u884c\r\n<\/span>    st<\/span>.<\/span>write<\/span>(<\/span>qa<\/span>.<\/span>run<\/span>(<\/span>input_text<\/span>))<\/span>\r\n<\/code><\/pre>\n
\u8bf7\u6309\u7167\u4ee5\u4e0b\u65b9\u5f0f\u6267\u884c\u3002<\/p>\n
python -m streamlit run pgv_search.py --server.port 8080\r\n<\/code><\/pre>\n
\"image.png\"<\/div>\n
\u901a\u8fc7\u542f\u7528Aurora Serverless PostgreSQL\u4e2d\u7684pgvector\uff0c\u6211\u4eec\u53ef\u4ee5\u5c06\u5176\u4f5c\u4e3aVector DB\uff08Vector\u5b58\u50a8\uff09\u6765\u4f7f\u7528\uff0c\u5e76\u5b9e\u73b0RAG\u3002<\/p>\n
\u8fd9\u4e00\u6b21LangChain\u5728\u5904\u7406\u4e0a\u505a\u5f97\u975e\u5e38\u597d\uff0c\u4f46\u662f\u5982\u679cAgents for Amazon Bedrock\u5f00\u59cb\u8fd0\u4f5c\uff0c\u5b83\u4eec\u4f1a\u4ee5\u540c\u6837\u826f\u597d\u7684\u65b9\u5f0f\u5904\u7406\u5417\uff1f<\/p>\n
 <\/p>\n","protected":false},"excerpt":{"rendered":"
\u6211\u4eec\u5c06\u4f7f\u7528Aurora Serverless PostgreSQL\u7684pgvector\u4f5c\u4e3aVector DB\u6765\u6784 […]<\/p>\n","protected":false},"author":5,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-42107","post","type-post","status-publish","format-standard","hentry","category-uncategorized"],"yoast_head":"\n\u4f7f\u7528Bedrock\u548cAurora Serverless PostgreSQL\u7684pgvector\u6765\u8fdb\u884cRAG\uff08Retrieve, Aggregate, Generate\uff09- \u5728Amazon Titan Embeddings\u4e2d\u5b66\u4e60\u5411\u91cf - Blog - Silicon Cloud<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.silicloud.com\/zh\/blog\/\u4f7f\u7528bedrock\u548caurora-serverless-postgresql\u7684pgvector\u6765\u8fdb\u884crag\uff08retrieve-aggregate-generate\uff09-\u5728amazon-titan-embeddings\u4e2d\/\" \/>\n<meta property=\"og:locale\" content=\"zh_CN\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"\u4f7f\u7528Bedrock\u548cAurora Serverless PostgreSQL\u7684pgvector\u6765\u8fdb\u884cRAG\uff08Retrieve, Aggregate, Generate\uff09- \u5728Amazon Titan Embeddings\u4e2d\u5b66\u4e60\u5411\u91cf\" \/>\n<meta property=\"og:description\" content=\"\u6211\u4eec\u5c06\u4f7f\u7528Aurora Serverless PostgreSQL\u7684pgvector\u4f5c\u4e3aVector DB\u6765\u6784 […]\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.silicloud.com\/zh\/blog\/\u4f7f\u7528bedrock\u548caurora-serverless-postgresql\u7684pgvector\u6765\u8fdb\u884crag\uff08retrieve-aggregate-generate\uff09-\u5728amazon-titan-embeddings\u4e2d\/\" \/>\n<meta property=\"og:site_name\" content=\"Blog - Silicon Cloud\" \/>\n<meta property=\"article:published_time\" content=\"2023-08-19T07:49:36+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2024-04-30T05:30:23+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/cdn.silicloud.com\/blog-img\/blog\/img\/657d4aa237434c4406cb3d96\/2-0.png\" \/>\n<meta name=\"author\" content=\"\u6e05, \u5b87\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u4f5c\u8005\" \/>\n\t<meta name=\"twitter:data1\" content=\"\u6e05, \u5b87\" \/>\n\t<meta name=\"twitter:label2\" content=\"\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4\" \/>\n\t<meta name=\"twitter:data2\" content=\"4 \u5206\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.silicloud.com\/zh\/blog\/%e4%bd%bf%e7%94%a8bedrock%e5%92%8caurora-serverless-postgresql%e7%9a%84pgvector%e6%9d%a5%e8%bf%9b%e8%a1%8crag%ef%bc%88retrieve-aggregate-generate%ef%bc%89-%e5%9c%a8amazon-titan-embeddings%e4%b8%ad\/\",\"url\":\"https:\/\/www.silicloud.com\/zh\/blog\/%e4%bd%bf%e7%94%a8bedrock%e5%92%8caurora-serverless-postgresql%e7%9a%84pgvector%e6%9d%a5%e8%bf%9b%e8%a1%8crag%ef%bc%88retrieve-aggregate-generate%ef%bc%89-%e5%9c%a8amazon-titan-embeddings%e4%b8%ad\/\",\"name\":\"\u4f7f\u7528Bedrock\u548cAurora Serverless PostgreSQL\u7684pgvector\u6765\u8fdb\u884cRAG\uff08Retrieve, Aggregate, Generate\uff09- \u5728Amazon Titan Embeddings\u4e2d\u5b66\u4e60\u5411\u91cf - Blog - Silicon Cloud\",\"isPartOf\":{\"@id\":\"https:\/\/www.silicloud.com\/zh\/blog\/#website\"},\"datePublished\":\"2023-08-19T07:49:36+00:00\",\"dateModified\":\"2024-04-30T05:30:23+00:00\",\"author\":{\"@id\":\"https:\/\/www.silicloud.com\/zh\/blog\/#\/schema\/person\/1a6ecd3d914d22a5ac32791ffc1fbd8e\"},\"breadcrumb\":{\"@id\":\"https:\/\/www.silicloud.com\/zh\/blog\/%e4%bd%bf%e7%94%a8bedrock%e5%92%8caurora-serverless-postgresql%e7%9a%84pgvector%e6%9d%a5%e8%bf%9b%e8%a1%8crag%ef%bc%88retrieve-aggregate-generate%ef%bc%89-%e5%9c%a8amazon-titan-embeddings%e4%b8%ad\/#breadcrumb\"},\"inLanguage\":\"zh-Hans\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.silicloud.com\/zh\/blog\/%e4%bd%bf%e7%94%a8bedrock%e5%92%8caurora-serverless-postgresql%e7%9a%84pgvector%e6%9d%a5%e8%bf%9b%e8%a1%8crag%ef%bc%88retrieve-aggregate-generate%ef%bc%89-%e5%9c%a8amazon-titan-embeddings%e4%b8%ad\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.silicloud.com\/zh\/blog\/%e4%bd%bf%e7%94%a8bedrock%e5%92%8caurora-serverless-postgresql%e7%9a%84pgvector%e6%9d%a5%e8%bf%9b%e8%a1%8crag%ef%bc%88retrieve-aggregate-generate%ef%bc%89-%e5%9c%a8amazon-titan-embeddings%e4%b8%ad\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"\u9996\u9875\",\"item\":\"https:\/\/www.silicloud.com\/zh\/blog\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"\u4f7f\u7528Bedrock\u548cAurora Serverless PostgreSQL\u7684pgvector\u6765\u8fdb\u884cRAG\uff08Retrieve, Aggregate, Generate\uff09- \u5728Amazon Titan Embeddings\u4e2d\u5b66\u4e60\u5411\u91cf\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.silicloud.com\/zh\/blog\/#website\",\"url\":\"https:\/\/www.silicloud.com\/zh\/blog\/\",\"name\":\"Blog - Silicon Cloud\",\"description\":\"\",\"inLanguage\":\"zh-Hans\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.silicloud.com\/zh\/blog\/#\/schema\/person\/1a6ecd3d914d22a5ac32791ffc1fbd8e\",\"name\":\"\u6e05, \u5b87\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/www.silicloud.com\/zh\/blog\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/secure.gravatar.com\/avatar\/4b2016c18459a605fc469c7566608f5686491baa112d0871ee613f61b7210565?s=96&d=mm&r=g\",\"contentUrl\":\"https:\/\/secure.gravatar.com\/avatar\/4b2016c18459a605fc469c7566608f5686491baa112d0871ee613f61b7210565?s=96&d=mm&r=g\",\"caption\":\"\u6e05, \u5b87\"},\"url\":\"https:\/\/www.silicloud.com\/zh\/blog\/author\/qingyu\/\"},{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/www.silicloud.com\/zh\/blog\/%e4%bd%bf%e7%94%a8bedrock%e5%92%8caurora-serverless-postgresql%e7%9a%84pgvector%e6%9d%a5%e8%bf%9b%e8%a1%8crag%ef%bc%88retrieve-aggregate-generate%ef%bc%89-%e5%9c%a8amazon-titan-embeddings%e4%b8%ad\/#local-main-organization-logo\",\"url\":\"\",\"contentUrl\":\"\",\"caption\":\"Blog - Silicon Cloud\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"\u4f7f\u7528Bedrock\u548cAurora Serverless PostgreSQL\u7684pgvector\u6765\u8fdb\u884cRAG\uff08Retrieve, Aggregate, Generate\uff09- \u5728Amazon Titan Embeddings\u4e2d\u5b66\u4e60\u5411\u91cf - Blog - Silicon Cloud","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.silicloud.com\/zh\/blog\/\u4f7f\u7528bedrock\u548caurora-serverless-postgresql\u7684pgvector\u6765\u8fdb\u884crag\uff08retrieve-aggregate-generate\uff09-\u5728amazon-titan-embeddings\u4e2d\/","og_locale":"zh_CN","og_type":"article","og_title":"\u4f7f\u7528Bedrock\u548cAurora Serverless PostgreSQL\u7684pgvector\u6765\u8fdb\u884cRAG\uff08Retrieve, Aggregate, Generate\uff09- \u5728Amazon Titan Embeddings\u4e2d\u5b66\u4e60\u5411\u91cf","og_description":"\u6211\u4eec\u5c06\u4f7f\u7528Aurora Serverless PostgreSQL\u7684pgvector\u4f5c\u4e3aVector DB\u6765\u6784 […]","og_url":"https:\/\/www.silicloud.com\/zh\/blog\/\u4f7f\u7528bedrock\u548caurora-serverless-postgresql\u7684pgvector\u6765\u8fdb\u884crag\uff08retrieve-aggregate-generate\uff09-\u5728amazon-titan-embeddings\u4e2d\/","og_site_name":"Blog - Silicon Cloud","article_published_time":"2023-08-19T07:49:36+00:00","article_modified_time":"2024-04-30T05:30:23+00:00","og_image":[{"url":"https:\/\/cdn.silicloud.com\/blog-img\/blog\/img\/657d4aa237434c4406cb3d96\/2-0.png"}],"author":"\u6e05, \u5b87","twitter_card":"summary_large_image","twitter_misc":{"\u4f5c\u8005":"\u6e05, \u5b87","\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4":"4 \u5206"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.silicloud.com\/zh\/blog\/%e4%bd%bf%e7%94%a8bedrock%e5%92%8caurora-serverless-postgresql%e7%9a%84pgvector%e6%9d%a5%e8%bf%9b%e8%a1%8crag%ef%bc%88retrieve-aggregate-generate%ef%bc%89-%e5%9c%a8amazon-titan-embeddings%e4%b8%ad\/","url":"https:\/\/www.silicloud.com\/zh\/blog\/%e4%bd%bf%e7%94%a8bedrock%e5%92%8caurora-serverless-postgresql%e7%9a%84pgvector%e6%9d%a5%e8%bf%9b%e8%a1%8crag%ef%bc%88retrieve-aggregate-generate%ef%bc%89-%e5%9c%a8amazon-titan-embeddings%e4%b8%ad\/","name":"\u4f7f\u7528Bedrock\u548cAurora Serverless PostgreSQL\u7684pgvector\u6765\u8fdb\u884cRAG\uff08Retrieve, Aggregate, Generate\uff09- \u5728Amazon Titan Embeddings\u4e2d\u5b66\u4e60\u5411\u91cf - Blog - Silicon Cloud","isPartOf":{"@id":"https:\/\/www.silicloud.com\/zh\/blog\/#website"},"datePublished":"2023-08-19T07:49:36+00:00","dateModified":"2024-04-30T05:30:23+00:00","author":{"@id":"https:\/\/www.silicloud.com\/zh\/blog\/#\/schema\/person\/1a6ecd3d914d22a5ac32791ffc1fbd8e"},"breadcrumb":{"@id":"https:\/\/www.silicloud.com\/zh\/blog\/%e4%bd%bf%e7%94%a8bedrock%e5%92%8caurora-serverless-postgresql%e7%9a%84pgvector%e6%9d%a5%e8%bf%9b%e8%a1%8crag%ef%bc%88retrieve-aggregate-generate%ef%bc%89-%e5%9c%a8amazon-titan-embeddings%e4%b8%ad\/#breadcrumb"},"inLanguage":"zh-Hans","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.silicloud.com\/zh\/blog\/%e4%bd%bf%e7%94%a8bedrock%e5%92%8caurora-serverless-postgresql%e7%9a%84pgvector%e6%9d%a5%e8%bf%9b%e8%a1%8crag%ef%bc%88retrieve-aggregate-generate%ef%bc%89-%e5%9c%a8amazon-titan-embeddings%e4%b8%ad\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/www.silicloud.com\/zh\/blog\/%e4%bd%bf%e7%94%a8bedrock%e5%92%8caurora-serverless-postgresql%e7%9a%84pgvector%e6%9d%a5%e8%bf%9b%e8%a1%8crag%ef%bc%88retrieve-aggregate-generate%ef%bc%89-%e5%9c%a8amazon-titan-embeddings%e4%b8%ad\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"\u9996\u9875","item":"https:\/\/www.silicloud.com\/zh\/blog\/"},{"@type":"ListItem","position":2,"name":"\u4f7f\u7528Bedrock\u548cAurora Serverless PostgreSQL\u7684pgvector\u6765\u8fdb\u884cRAG\uff08Retrieve, Aggregate, Generate\uff09- \u5728Amazon Titan Embeddings\u4e2d\u5b66\u4e60\u5411\u91cf"}]},{"@type":"WebSite","@id":"https:\/\/www.silicloud.com\/zh\/blog\/#website","url":"https:\/\/www.silicloud.com\/zh\/blog\/","name":"Blog - Silicon Cloud","description":"","inLanguage":"zh-Hans"},{"@type":"Person","@id":"https:\/\/www.silicloud.com\/zh\/blog\/#\/schema\/person\/1a6ecd3d914d22a5ac32791ffc1fbd8e","name":"\u6e05, \u5b87","image":{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/www.silicloud.com\/zh\/blog\/#\/schema\/person\/image\/","url":"https:\/\/secure.gravatar.com\/avatar\/4b2016c18459a605fc469c7566608f5686491baa112d0871ee613f61b7210565?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/4b2016c18459a605fc469c7566608f5686491baa112d0871ee613f61b7210565?s=96&d=mm&r=g","caption":"\u6e05, \u5b87"},"url":"https:\/\/www.silicloud.com\/zh\/blog\/author\/qingyu\/"},{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/www.silicloud.com\/zh\/blog\/%e4%bd%bf%e7%94%a8bedrock%e5%92%8caurora-serverless-postgresql%e7%9a%84pgvector%e6%9d%a5%e8%bf%9b%e8%a1%8crag%ef%bc%88retrieve-aggregate-generate%ef%bc%89-%e5%9c%a8amazon-titan-embeddings%e4%b8%ad\/#local-main-organization-logo","url":"","contentUrl":"","caption":"Blog - Silicon Cloud"}]}},"_links":{"self":[{"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/posts\/42107","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/users\/5"}],"replies":[{"embeddable":true,"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/comments?post=42107"}],"version-history":[{"count":2,"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/posts\/42107\/revisions"}],"predecessor-version":[{"id":92195,"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/posts\/42107\/revisions\/92195"}],"wp:attachment":[{"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/media?parent=42107"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/categories?post=42107"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/tags?post=42107"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}