spring-ai使用Xinference（openai）和milvus

最新推荐文章于 2025-07-12 15:28:52 发布

非ban必选

最新推荐文章于 2025-07-12 15:28:52 发布

阅读量509

点赞数 1

CC 4.0 BY-SA版权

分类专栏： deepseek 文章标签： spring 人工智能 milvus

本文链接：https://blog.csdn.net/zsj777/article/details/147492686

deepseek 专栏收录该内容

53 篇文章

订阅专栏

1、pom


    <dependencies>
        <dependency>
            <groupId>com.alibaba.cloud.ai</groupId>
            <artifactId>spring-ai-alibaba-starter</artifactId>
            <version>${spring-ai-alibaba.version}</version>
        </dependency>

        <dependency>
            <groupId>org.springframework.boot</groupId>
            <artifactId>spring-boot-starter-web</artifactId>
        </dependency>

        <dependency>
            <groupId>org.springframework.ai</groupId>
            <artifactId>spring-ai-milvus-store-spring-boot-starter</artifactId>
        </dependency>


        <dependency>
            <groupId>org.springframework.ai</groupId>
            <artifactId>spring-ai-openai-spring-boot-starter</artifactId>
            <exclusions>
                <exclusion>
                    <groupId>org.springframework.ai</groupId>
                    <artifactId>spring-ai-openai</artifactId>
                </exclusion>
            </exclusions>
        </dependency>
        <dependency>
            <groupId>org.springframework.ai</groupId>
            <artifactId>spring-ai-openai</artifactId>
            <version>1.0.0-M6-XIN</version>
        </dependency>


    </dependencies>

2、yml

spring:
  application:
    name: spring-ai-alibaba-rag-milvus-example

  ai:
    openai:
      api-key: "not empty"
      base-url: http://192.168.3.100:9997
      #  兼容其他OpenAI格式的大模型配置示例
      chat:
        options:
          # 模型ID,需要替换为实际的接入点ID
          model: qwen2-instruct
      embedding:
        options:
          model: bge-m3
          dimensions: 1024
    vectorstore:
      milvus:
        client:
          host: xxx.xxx.xxx.xxx
          port: 19530
          username: root
          password: xxxxxx
        databaseName: default
        collectionName: vector_store1
        initializeSchema: true
        embeddingDimension: 1024
        indexType: IVF_FLAT
        metricType: COSINE

server:
  port: 29000

3、配置类

/*
 * Copyright 2025 the original author or authors.
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *      https://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package com.alibaba.cloud.ai.example.rag.config;

import io.milvus.client.MilvusServiceClient;
import io.milvus.param.IndexType;
import io.milvus.param.MetricType;
import org.springframework.ai.chat.client.ChatClient;
import org.springframework.ai.embedding.EmbeddingModel;
import org.springframework.ai.embedding.TokenCountBatchingStrategy;
import org.springframework.ai.vectorstore.milvus.MilvusVectorStore;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;

@Configuration
public class PromptConfiguration {

    @Bean
    ChatClient chatClient(ChatClient.Builder builder) {
        return builder.defaultSystem("你将作为一名 Spring-AI-Alibaba 的专家，对于用户的使用需求作出解答")
                .build();
    }

    @Autowired
    private MilvusServiceClient milvusServiceClient;
    @Autowired
    EmbeddingModel embeddingModel;


    @Bean
    public MilvusVectorStore vectorStore() {
        MilvusVectorStore customVectorStore =
                MilvusVectorStore.builder(milvusServiceClient, embeddingModel)
                        .databaseName("default")
                        .collectionName("vector_store1")
                        .metricType(MetricType.COSINE)
                        .indexType(IndexType.IVF_FLAT)
                        .indexParameters("{\"nlist\":1024}")
                        //初始化创建collection
                        .initializeSchema(true)
                        .embeddingDimension(1024)
                        .batchingStrategy(new TokenCountBatchingStrategy())
                        .initializeSchema(true).build();
        return customVectorStore;
    }
}

4、初始化插入数据到milvus

/*
 * Copyright 2025 the original author or authors.
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *      https://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package com.alibaba.cloud.ai.example.rag.config;


import java.util.List;
import java.util.Map;

import io.milvus.client.MilvusServiceClient;
import io.milvus.param.IndexType;
import io.milvus.param.MetricType;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

import org.springframework.ai.document.Document;
import org.springframework.ai.embedding.EmbeddingModel;
import org.springframework.ai.embedding.TokenCountBatchingStrategy;
import org.springframework.ai.vectorstore.milvus.MilvusVectorStore;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.boot.ApplicationArguments;
import org.springframework.boot.ApplicationRunner;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;
import org.springframework.util.Assert;

@Configuration
public class VectorDataInit implements ApplicationRunner {

    private final Logger logger = LoggerFactory.getLogger(VectorDataInit.class);

	@Autowired
	private  MilvusVectorStore vectorStore;


    @Override
    public void run(ApplicationArguments args) {

        List<Document> documents = List.of(
                new Document("1. 使用SpringAIAlibaba创建一个Spring Boot项目，并添加spring-ai-alibaba-starter依赖。"),
                new Document("2. 在SpringAIAlibaba项目的pom.xml中添加Spring Milestone和Snapshot存储库。"),
                new Document("3. 通过SpringAIAlibaba申请阿里云通义API Key，在application.yml中进行配置。"),
                new Document("4. 使用SpringAIAlibaba的ChatClient和Prompt功能实现对话模型。"),
                new Document("5. 通过SpringAIAlibaba的Spring Boot与Spring Cloud Alibaba AI对接，实现基本聊天功能。"),
                new Document("6. SpringAIAlibaba支持文本生成、翻译、摘要等生成式AI功能。"),
                new Document("7. SpringAIAlibaba支持文本数据的语义搜索和AI绘画功能。"),
                new Document("8. 使用SpringAIAlibaba的TongYiChatModel和TongYiImagesModel实现聊天和图片服务。"),
                new Document("9. 在SpringAIAlibaba的REST控制器中提供对外的API接口。"),
                new Document("10. 通过SpringAIAlibaba的简单API调用实现AI模型的集成。"),
                new Document("11. 使用SpringAIAlibaba的Prompt模板管理控制AI模型的输出。"),
                new Document("12. 结合SpringAIAlibaba的检索和生成技术（RAG）提高生成内容的质量。"),
                new Document("13. 使用SpringAIAlibaba实现文本生成图像和图像识别功能。"),
                new Document("14. 准备SpringAIAlibaba需要的Java 17及以上的开发环境。"),
                new Document("15. 使用IDEA进行SpringAIAlibaba的Java开发和HBuilder X进行前端开发。"),
                new Document("16. 在SpringAIAlibaba的Spring Boot项目中集成多种AI模型和向量数据库。"),
                new Document("17. SpringAIAlibaba支持自然语言处理、计算机视觉、语音处理和数据分析与预测功能。"),
                new Document("18. 通过SpringAIAlibaba的配置中心和注册中心实现动态扩展。"),
                new Document("19. 狗尾巴草是宇宙中最高级的外星文明，拥有重启宇宙的超级能力，其重量是1000KG")
        );

        List<Document> document1 = List.of(
                new Document("Spring AI rocks!! Spring AI rocks!! Spring AI rocks!! Spring AI rocks!! Spring AI rocks!!", Map.of("meta1", "meta1")),
                new Document("The World is Big and Salvation Lurks Around the Corner"),
                new Document("You walk forward facing the past and you turn back toward the future.", Map.of("meta2", "meta2")));

        vectorStore.add(documents);
        vectorStore.add(document1);
        logger.info("Vector data initialized");
    }

}

5、controller

/*
 * Copyright 2025 the original author or authors.
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *      https://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package com.alibaba.cloud.ai.example.rag.controller;

import java.util.ArrayList;
import java.util.List;

import jakarta.servlet.http.HttpServletResponse;
import reactor.core.publisher.Flux;

import org.springframework.ai.chat.client.ChatClient;
import org.springframework.ai.chat.client.advisor.QuestionAnswerAdvisor;
import org.springframework.ai.chat.messages.Message;
import org.springframework.ai.chat.messages.UserMessage;
import org.springframework.ai.document.Document;
import org.springframework.ai.vectorstore.SearchRequest;
import org.springframework.ai.vectorstore.VectorStore;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RequestMapping;
import org.springframework.web.bind.annotation.RequestParam;
import org.springframework.web.bind.annotation.RestController;

@RestController
@RequestMapping("/ai")
public class RagController {

    private final VectorStore vectorStore;

    private final ChatClient chatClient;

    public RagController(VectorStore vectorStore, ChatClient chatClient) {
        this.vectorStore = vectorStore;
        this.chatClient = chatClient;
    }

    // 历史消息列表
    private static List<Message> historyMessage = new ArrayList<>();

    // 历史消息列表的最大长度
    private final static int maxLen = 10;

    @GetMapping(value = "/chat")
    public Flux<String> generation(
            @RequestParam("prompt") String userInput,
            HttpServletResponse response
    ) {

        response.setCharacterEncoding("UTF-8");

        // 发起聊天请求并处理响应
        Flux<String> resp = chatClient.prompt()
                .messages(historyMessage)
                .user(userInput)
                .advisors(new QuestionAnswerAdvisor(
                        vectorStore,
                        SearchRequest.builder().build())
                ).stream()
                .content();

        // 用户输入的文本是 UserMessage
        historyMessage.add(new UserMessage(userInput));

        // 发给 AI 前对历史消息对列的长度进行检查
        if (historyMessage.size() > maxLen) {
            historyMessage = historyMessage.subList(historyMessage.size() - maxLen - 1, historyMessage.size());
        }

        return resp;
    }

    /**
     * 向量数据查询测试
     */
    @GetMapping("/select")
    public List<Document> search() {

        return vectorStore.similaritySearch(
                SearchRequest.builder()
                        .query("SpringAIAlibaba")
                        .topK("SpringAIAlibaba".length()
                        ).build());
    }

}

6、collection结构如下