按语义相似性搜索向量存储文档

通过语义相似性和批量操作搜索向量存储文档

get

Initiate a search operation with a query text and receive most semantically similar responses from the vector store.

Path parameters

vector-store-idstringRequired

The ID of the vector store.

Query parameters

qstringRequired

The search query text.

nintegerOptional

Number of chunks to return.

tnumberOptional

Matching threshold.

prev_chunksintegerOptional

Number of chunks before the matched chunk to include.

next_chunksintegerOptional

Number of chunks after the matched chunk to include.

Header parameters

x-api-keystringRequired

The API key for authentication.

Responses

200

Search results retrieved successfully.

application/json

get

/vector-stores/{vector-store-id}/documents/search

GET /api/v1/vector-stores/{vector-store-id}/documents/search?q=text HTTP/1.1
Host: api.rememberizer.ai
x-api-key: text
Accept: */*

200

Search results retrieved successfully.

{
  "vector_store": {
    "id": "text",
    "name": "text"
  },
  "matched_chunks": [
    {
      "document": {
        "id": 1,
        "name": "text",
        "type": "text",
        "size": 1,
        "indexed_on": "2025-12-27T13:31:01.504Z",
        "vector_store": "text",
        "created": "2025-12-27T13:31:01.504Z",
        "modified": "2025-12-27T13:31:01.504Z"
      },
      "matched_content": "text",
      "distance": 1
    }
  ]
}

示例请求

curl -X GET \
  "https://api.rememberizer.ai/api/v1/vector-stores/vs_abc123/documents/search?q=如何将我们的产品与第三方系统集成&n=5&prev_chunks=1&next_chunks=1" \
  -H "x-api-key: YOUR_API_KEY"

将 YOUR_API_KEY 替换为您的实际向量存储 API 密钥，将 vs_abc123 替换为您的向量存储 ID。

const searchVectorStore = async (vectorStoreId, query, numResults = 5, prevChunks = 1, nextChunks = 1) => {
  const url = new URL(`https://api.rememberizer.ai/api/v1/vector-stores/${vectorStoreId}/documents/search`);
  url.searchParams.append('q', query);
  url.searchParams.append('n', numResults);
  url.searchParams.append('prev_chunks', prevChunks);
  url.searchParams.append('next_chunks', nextChunks);
  
  const response = await fetch(url.toString(), {
    method: 'GET',
    headers: {
      'x-api-key': 'YOUR_API_KEY'
    }
  });
  
  const data = await response.json();
  console.log(data);
};

searchVectorStore(
  'vs_abc123',
  '如何将我们的产品与第三方系统集成',
  5,
  1,
  1
);

将 YOUR_API_KEY 替换为您的实际向量存储 API 密钥，将 vs_abc123 替换为您的向量存储 ID。

import requests

def search_vector_store(vector_store_id, query, num_results=5, prev_chunks=1, next_chunks=1):
    headers = {
        "x-api-key": "YOUR_API_KEY"
    }
    
    params = {
        "q": query,
        "n": num_results,
        "prev_chunks": prev_chunks,
        "next_chunks": next_chunks
    }
    
    response = requests.get(
        f"https://api.rememberizer.ai/api/v1/vector-stores/{vector_store_id}/documents/search",
        headers=headers,
        params=params
    )
    
    data = response.json()
    print(data)

search_vector_store(
    'vs_abc123',
    '如何将我们的产品与第三方系统集成',
    5,
    1,
    1
)

将 YOUR_API_KEY 替换为您的实际向量存储 API 密钥，将 vs_abc123 替换为您的向量存储 ID。

require 'net/http'
require 'uri'
require 'json'

def search_vector_store(vector_store_id, query, num_results=5, prev_chunks=1, next_chunks=1)
  uri = URI("https://api.rememberizer.ai/api/v1/vector-stores/#{vector_store_id}/documents/search")
  params = {
    q: query,
    n: num_results,
    prev_chunks: prev_chunks,
    next_chunks: next_chunks
  }
  
  uri.query = URI.encode_www_form(params)
  
  request = Net::HTTP::Get.new(uri)
  request['x-api-key'] = 'YOUR_API_KEY'
  
  http = Net::HTTP.new(uri.host, uri.port)
  http.use_ssl = true
  
  response = http.request(request)
  data = JSON.parse(response.body)
  puts data
end

search_vector_store(
  'vs_abc123',
  '如何将我们的产品与第三方系统集成',
  5,
  1,
  1
)

将 YOUR_API_KEY 替换为您的实际向量存储 API 密钥，将 vs_abc123 替换为您的向量存储 ID。

路径参数

参数

类型

描述

vector-store-id

字符串

必填。 要搜索的向量存储的 ID。

查询参数

参数

类型

描述

字符串

必填。 搜索查询文本。

整数

返回的结果数量。默认：10。

数字

匹配阈值。默认：0.7。

prev_chunks

整数

包含匹配块之前的块数。默认：0。

next_chunks

整数

包含匹配块之后的块数。默认：0。

响应格式

{
  "vector_store": {
    "id": "vs_abc123",
    "name": "产品文档"
  },
  "matched_chunks": [
    {
      "document": {
        "id": 1234,
        "name": "集成指南.pdf",
        "type": "application/pdf",
        "size": 250000,
        "indexed_on": "2023-06-15T10:30:00Z",
        "vector_store": "vs_abc123",
        "created": "2023-06-15T10:15:00Z",
        "modified": "2023-06-15T10:30:00Z"
      },
      "matched_content": "我们的产品为第三方系统提供了几种集成选项。主要方法是通过我们的 RESTful API，支持 OAuth2 认证。此外，您还可以使用我们提供的 Python、JavaScript 和 Java 的 SDK。",
      "distance": 0.123
    },
    // ... 更多匹配的内容
  ]
}

认证

此端点需要使用 x-api-key 头中的 API 密钥进行认证。

错误响应

状态码

描述

400

错误请求 - 缺少必需的参数或格式无效

401

未授权 - API 密钥无效或缺失

404

未找到 - 找不到向量存储

500

服务器内部错误

搜索优化技巧

上下文窗口

使用 prev_chunks 和 next_chunks 参数来控制每个匹配包含多少上下文：

将两者都设置为 0 以获得没有上下文的精确匹配
将两者都设置为 1-2 以获得具有最小上下文的匹配
将两者都设置为 3-5 以获得具有大量上下文的匹配

匹配阈值

t 参数控制匹配的过滤严格程度：

较高的值（例如，0.9）仅返回非常接近的匹配
较低的值（例如，0.5）返回更多种类的匹配
默认值（0.7）提供了一种平衡的方法

批量操作

对于高吞吐量的应用，Rememberizer 支持在向量存储上进行高效的批量操作。这些方法在处理多个搜索查询时优化性能。

批量搜索实现

import requests
import time
import concurrent.futures

def batch_search_vector_store(vector_store_id, queries, num_results=5, batch_size=10):
    """
    对向量存储执行批量搜索
    
    参数:
        vector_store_id: 要搜索的向量存储的ID
        queries: 搜索查询字符串列表
        num_results: 每个查询的结果数量
        batch_size: 并行请求的数量
        
    返回:
        搜索结果列表
    """
    headers = {
        "x-api-key": "YOUR_API_KEY"
    }
    
    results = []
    
    # 分批处理以避免过载API
    for i in range(0, len(queries), batch_size):
        batch_queries = queries[i:i+batch_size]
        
        with concurrent.futures.ThreadPoolExecutor(max_workers=batch_size) as executor:
            futures = []
            
            for query in batch_queries:
                params = {
                    "q": query,
                    "n": num_results,
                    "prev_chunks": 1,
                    "next_chunks": 1
                }
                
                # 提交请求到线程池
                future = executor.submit(
                    requests.get,
                    f"https://api.rememberizer.ai/api/v1/vector-stores/{vector_store_id}/documents/search",
                    headers=headers,
                    params=params
                )
                futures.append(future)
            
            # 从所有未来对象收集结果
            for future in futures:
                response = future.result()
                if response.status_code == 200:
                    results.append(response.json())
                else:
                    results.append({"error": f"状态码失败: {response.status_code}"})
        
        # 在批次之间添加延迟以避免速率限制
        if i + batch_size < len(queries):
            time.sleep(1)
    
    return results

# 示例用法
queries = [
    "与 REST API 的集成",
    "认证协议",
    "如何部署到生产环境",
    "性能优化技术",
    "错误处理最佳实践"
]

search_results = batch_search_vector_store("vs_abc123", queries, num_results=3, batch_size=5)

/**
 * 对向量存储执行批量搜索
 * 
 * @param {string} vectorStoreId - 向量存储的 ID
 * @param {string[]} queries - 搜索查询列表
 * @param {Object} options - 配置选项
 * @returns {Promise<Array>} - 搜索结果列表
 */
async function batchSearchVectorStore(vectorStoreId, queries, options = {}) {
  const {
    numResults = 5,
    batchSize = 10,
    delayBetweenBatches = 1000,
    prevChunks = 1,
    nextChunks = 1,
    distanceThreshold = 0.7
  } = options;
  
  const results = [];
  const apiKey = 'YOUR_API_KEY';
  
  // 以批次处理以管理 API 负载
  for (let i = 0; i < queries.length; i += batchSize) {
    const batchQueries = queries.slice(i, i + batchSize);
    
    // 创建并行请求的 promise 数组
    const batchPromises = batchQueries.map(query => {
      const url = new URL(`https://api.rememberizer.ai/api/v1/vector-stores/${vectorStoreId}/documents/search`);
      url.searchParams.append('q', query);
      url.searchParams.append('n', numResults);
      url.searchParams.append('prev_chunks', prevChunks);
      url.searchParams.append('next_chunks', nextChunks);
      url.searchParams.append('t', distanceThreshold);
      
      return fetch(url.toString(), {
        method: 'GET',
        headers: {
          'x-api-key': apiKey
        }
      })
      .then(response => {
        if (response.ok) {
          return response.json();
        } else {
          return { error: `失败，状态码: ${response.status}` };
        }
      })
      .catch(error => {
        return { error: error.message };
      });
    });
    
    // 等待批次中的所有请求完成
    const batchResults = await Promise.all(batchPromises);
    results.push(...batchResults);
    
    // 在批次之间添加延迟以避免速率限制
    if (i + batchSize < queries.length) {
      await new Promise(resolve => setTimeout(resolve, delayBetweenBatches));
    }
  }
  
  return results;
}

// 示例用法
const queries = [
  "与 REST API 的集成",
  "认证协议",
  "如何部署到生产环境",
  "性能优化技术",
  "错误处理最佳实践"
];

const options = {
  numResults: 3,
  batchSize: 5,
  delayBetweenBatches: 1000,
  prevChunks: 1,
  nextChunks: 1
};

batchSearchVectorStore("vs_abc123", queries, options)
  .then(results => console.log(results))
  .catch(error => console.error("批量搜索失败:", error));

require 'net/http'
require 'uri'
require 'json'
require 'concurrent'

# 对向量存储执行批量搜索
#
# @param vector_store_id [String] 向量存储的 ID
# @param queries [Array<String>] 搜索查询列表
# @param num_results [Integer] 每个查询的结果数量
# @param batch_size [Integer] 并行请求的数量
# @param delay_between_batches [Float] 批次之间等待的秒数
# @return [Array] 每个查询的搜索结果
def batch_search_vector_store(vector_store_id, queries, num_results: 5, batch_size: 10, delay_between_batches: 1.0)
  results = []
  api_key = 'YOUR_API_KEY'
  
  # 按批处理
  queries.each_slice(batch_size).with_index do |batch_queries, batch_index|
    # 创建一个线程池以进行并发执行
    pool = Concurrent::FixedThreadPool.new(batch_size)
    futures = []
    
    batch_queries.each do |query|
      # 将每个请求提交到线程池
      futures << Concurrent::Future.execute(executor: pool) do
        uri = URI("https://api.rememberizer.ai/api/v1/vector-stores/#{vector_store_id}/documents/search")
        params = {
          q: query,
          n: num_results,
          prev_chunks: 1,
          next_chunks: 1
        }
        
        uri.query = URI.encode_www_form(params)
        
        request = Net::HTTP::Get.new(uri)
        request['x-api-key'] = api_key
        
        http = Net::HTTP.new(uri.host, uri.port)
        http.use_ssl = true
        
        begin
          response = http.request(request)
          
          if response.code.to_i == 200
            JSON.parse(response.body)
          else
            { "error" => "状态码失败: #{response.code}" }
          end
        rescue => e
          { "error" => e.message }
        end
      end
    end
    
    # 收集所有 futures 的结果
    batch_results = futures.map(&:value)
    results.concat(batch_results)
    
    # 在批次之间添加延迟
    if batch_index < (queries.length / batch_size.to_f).ceil - 1
      sleep(delay_between_batches)
    end
  end
  
  pool.shutdown
  results
end

# 示例用法
queries = [
  "与 REST API 的集成",
  "身份验证协议",
  "如何部署到生产环境", 
  "性能优化技术",
  "错误处理最佳实践"
]

results = batch_search_vector_store(
  "vs_abc123", 
  queries, 
  num_results: 3, 
  batch_size: 5
)

puts results

批量操作的性能优化

在实现向量存储搜索的批量操作时，请考虑以下最佳实践：

最佳批量大小：对于大多数应用程序，平行处理 5-10 个查询在吞吐量和资源使用之间提供了良好的平衡。
速率限制意识：在批次之间包含延迟机制（通常为 1-2 秒），以避免触及 API 速率限制。
错误处理：为可能在批次中失败的单个查询实现强大的错误处理。
连接管理：对于高流量应用程序，实施连接池以减少开销。
超时配置：为每个请求设置适当的超时，以防止长时间运行的查询阻塞整个批次。
结果处理：考虑在结果可用时异步处理结果，而不是等待所有结果。
监控：跟踪平均响应时间和成功率等性能指标，以识别优化机会。

对于查询量非常大的生产应用程序，考虑实施一个队列系统和工作进程，以高效管理大批量。

此端点允许您使用语义相似性搜索您的向量存储。它返回与您的查询在概念上相关的文档，即使它们不包含确切的关键字。这使得它在自然语言查询和问答中尤其强大。

Previous在向量存储中删除文档 Next更新向量存储中文件的内容

Last updated 1 month ago

import requests import time import concurrent.futures def batch_search_vector_store(vector_store_id, queries, num_results=5, batch_size=10): """ 对向量存储执行批量搜索参数: vector_store_id: 要搜索的向量存储的ID queries: 搜索查询字符串列表 num_results: 每个查询的结果数量 batch_size: 并行请求的数量返回: 搜索结果列表 """ headers = { "x-api-key": "YOUR_API_KEY" } results = [] # 分批处理以避免过载API for i in range(0, len(queries), batch_size): batch_queries = queries[i:i+batch_size] with concurrent.futures.ThreadPoolExecutor(max_workers=batch_size) as executor: futures = [] for query in batch_queries: params = { "q": query, "n": num_results, "prev_chunks": 1, "next_chunks": 1 } # 提交请求到线程池 future = executor.submit( requests.get, f"https://api.rememberizer.ai/api/v1/vector-stores/{vector_store_id}/documents/search", headers=headers, params=params ) futures.append(future) # 从所有未来对象收集结果 for future in futures: response = future.result() if response.status_code == 200: results.append(response.json()) else: results.append({"error": f"状态码失败: {response.status_code}"}) # 在批次之间添加延迟以避免速率限制 if i + batch_size < len(queries): time.sleep(1) return results # 示例用法 queries = [ "与 REST API 的集成", "认证协议", "如何部署到生产环境", "性能优化技术", "错误处理最佳实践" ] search_results = batch_search_vector_store("vs_abc123", queries, num_results=3, batch_size=5)

/** * 对向量存储执行批量搜索 * * @param {string} vectorStoreId - 向量存储的 ID * @param {string[]} queries - 搜索查询列表 * @param {Object} options - 配置选项 * @returns {Promise<Array>} - 搜索结果列表 */ async function batchSearchVectorStore(vectorStoreId, queries, options = {}) { const { numResults = 5, batchSize = 10, delayBetweenBatches = 1000, prevChunks = 1, nextChunks = 1, distanceThreshold = 0.7 } = options; const results = []; const apiKey = 'YOUR_API_KEY'; // 以批次处理以管理 API 负载 for (let i = 0; i < queries.length; i += batchSize) { const batchQueries = queries.slice(i, i + batchSize); // 创建并行请求的 promise 数组 const batchPromises = batchQueries.map(query => { const url = new URL(`https://api.rememberizer.ai/api/v1/vector-stores/${vectorStoreId}/documents/search`); url.searchParams.append('q', query); url.searchParams.append('n', numResults); url.searchParams.append('prev_chunks', prevChunks); url.searchParams.append('next_chunks', nextChunks); url.searchParams.append('t', distanceThreshold); return fetch(url.toString(), { method: 'GET', headers: { 'x-api-key': apiKey } }) .then(response => { if (response.ok) { return response.json(); } else { return { error: `失败，状态码: ${response.status}` }; } }) .catch(error => { return { error: error.message }; }); }); // 等待批次中的所有请求完成 const batchResults = await Promise.all(batchPromises); results.push(...batchResults); // 在批次之间添加延迟以避免速率限制 if (i + batchSize < queries.length) { await new Promise(resolve => setTimeout(resolve, delayBetweenBatches)); } } return results; } // 示例用法 const queries = [ "与 REST API 的集成", "认证协议", "如何部署到生产环境", "性能优化技术", "错误处理最佳实践" ]; const options = { numResults: 3, batchSize: 5, delayBetweenBatches: 1000, prevChunks: 1, nextChunks: 1 }; batchSearchVectorStore("vs_abc123", queries, options) .then(results => console.log(results)) .catch(error => console.error("批量搜索失败:", error));

require 'net/http' require 'uri' require 'json' require 'concurrent' # 对向量存储执行批量搜索 # # @param vector_store_id [String] 向量存储的 ID # @param queries [Array<String>] 搜索查询列表 # @param num_results [Integer] 每个查询的结果数量 # @param batch_size [Integer] 并行请求的数量 # @param delay_between_batches [Float] 批次之间等待的秒数 # @return [Array] 每个查询的搜索结果 def batch_search_vector_store(vector_store_id, queries, num_results: 5, batch_size: 10, delay_between_batches: 1.0) results = [] api_key = 'YOUR_API_KEY' # 按批处理 queries.each_slice(batch_size).with_index do |batch_queries, batch_index| # 创建一个线程池以进行并发执行 pool = Concurrent::FixedThreadPool.new(batch_size) futures = [] batch_queries.each do |query| # 将每个请求提交到线程池 futures << Concurrent::Future.execute(executor: pool) do uri = URI("https://api.rememberizer.ai/api/v1/vector-stores/#{vector_store_id}/documents/search") params = { q: query, n: num_results, prev_chunks: 1, next_chunks: 1 } uri.query = URI.encode_www_form(params) request = Net::HTTP::Get.new(uri) request['x-api-key'] = api_key http = Net::HTTP.new(uri.host, uri.port) http.use_ssl = true begin response = http.request(request) if response.code.to_i == 200 JSON.parse(response.body) else { "error" => "状态码失败: #{response.code}" } end rescue => e { "error" => e.message } end end end # 收集所有 futures 的结果 batch_results = futures.map(&:value) results.concat(batch_results) # 在批次之间添加延迟 if batch_index < (queries.length / batch_size.to_f).ceil - 1 sleep(delay_between_batches) end end pool.shutdown results end # 示例用法 queries = [ "与 REST API 的集成", "身份验证协议", "如何部署到生产环境", "性能优化技术", "错误处理最佳实践" ] results = batch_search_vector_store( "vs_abc123", queries, num_results: 3, batch_size: 5 ) puts results