langchain4j基础实战

快乐男孩30
304 阅读19分钟

langchain4j基础实战

1. 简介

langchain for java，在java语言中驾驭大模型的框架。langchain4j的目标是简化将LLM集成到java应用程序中的过程。

官网：docs.langchain4j.dev/

github：github.com/langchain4j…

image

2. 基础使用

2.1. helloworld

2.1.1. 前置

接入阿里云百炼平台的通义千问模型。bailian.console.aliyun.com

image

调用大模型之前我们需要准备api-key、模型名称、baseURL。

api-key：sk-c****3b12

image

模型名称：qwen-plusimage

baseUrl：dashscope.aliyuncs.com/compatible-…

image

引入依赖

image

<dependency>
    <groupId>dev.langchain4j</groupId>
    <artifactId>langchain4j-open-ai</artifactId>
</dependency>
<!--langchain4j 高阶-->
<dependency>
    <groupId>dev.langchain4j</groupId>
    <artifactId>langchain4j</artifactId>
</dependency>

2.1.2. 代码

写yaml

server.port=9001

spring.application.name=langchain4j-01helloworld

主启动类

import org.springframework.boot.SpringApplication;
import org.springframework.boot.autoconfigure.SpringBootApplication;

@SpringBootApplication
public class HelloLangChain4JApp
{
    public static void main(String[] args)
    {
        SpringApplication.run(HelloLangChain4JApp.class,args);
    }
}

配置类

import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.openai.OpenAiChatModel;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;


@Configuration
public class LLMConfig
{
    @Bean
    public ChatModel chatModelQwen()
    {
        return OpenAiChatModel.builder()
//                        .apiKey(System.getenv("aliQwen-api"))  //配置在环境变量中
                        .apiKey("sk-cf4de7068032411b9de71adde80e3b12")
                        .modelName("qwen-plus")
                        .baseUrl("https://dashscope.aliyuncs.com/compatible-mode/v1")
                .build();
    }

}

controller

import dev.langchain4j.data.message.UserMessage;
import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.openai.OpenAiChatModel;
import jakarta.annotation.Resource;
import lombok.extern.slf4j.Slf4j;
import org.apache.tomcat.util.buf.Utf8Encoder;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RequestParam;
import org.springframework.web.bind.annotation.RestController;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-27 21:43
 * @Description: TODO
 */
@RestController
@Slf4j
public class HelloLangChain4JController
{

    @Resource
    private ChatModel chatModel;

    @GetMapping(value = "/langchain4j/hello")
    public String hello(@RequestParam(value = "question",defaultValue = "你是谁") String question)
    {
        String result = chatModel.chat(question);

        System.out.println("调用大模型回复: "+result);

        return result;
    }

    public void test1(String question)
    {

    }
}

2.1.3. 测试

http://localhost:9001/langchain4j/hello?question=如何学习Java

image

2.2. 多模型共存

使用qwen-plus和DeepSeek模型共存

2.2.1. 准备

api-key：sk-fbe1701f86464a11bfa59b8dfbcf7f69

image

模型名称：deepseek-chat

base URL：api.deepseek.com/v1

image

2.2.2. 代码

配置类：

import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.openai.OpenAiChatModel;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;

@Configuration
public class LLMConfig
{
    @Bean(name = "qwen")
    public ChatModel chatModelQwen()
    {
        return OpenAiChatModel.builder()
                    .apiKey("sk-cf4de7068032411b9de71adde80e3b12")
                    .modelName("qwen-plus")
                    .baseUrl("https://dashscope.aliyuncs.com/compatible-mode/v1")
                .build();
    }

    @Bean(name = "deepseek")
    public ChatModel chatModelDeepSeek()
    {
        return
                OpenAiChatModel.builder()
                        .apiKey("sk-fbe1701f86464a11bfa59b8dfbcf7f69")
                        .modelName("deepseek-chat")
                        //.modelName("deepseek-reasoner")
                        .baseUrl("https://api.deepseek.com/v1")
                .build();
    }
}

controller：

import dev.langchain4j.model.chat.ChatModel;
import jakarta.annotation.Resource;
import lombok.extern.slf4j.Slf4j;
import org.springframework.boot.autoconfigure.SpringBootApplication;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RequestParam;
import org.springframework.web.bind.annotation.RestController;

@RestController
@Slf4j
public class MultiModelController
{
    @Resource(name = "qwen")
    private ChatModel chatModelQwen;

    @Resource(name = "deepseek")
    private ChatModel chatModelDeepSeek;

    // http://localhost:9002/multimodel/qwen
    @GetMapping(value = "/multimodel/qwen")
    public String qwenCall(@RequestParam(value = "prompt", defaultValue = "你是谁") String prompt)
    {
        String result = chatModelQwen.chat(prompt);

        System.out.println("通过langchain4j调用模型返回结果：\n"+result);

        return result;
    }

    // http://localhost:9002/multimodel/deepseek
    @GetMapping(value = "/multimodel/deepseek")
    public String deepseekCall(@RequestParam(value = "prompt", defaultValue = "你是谁") String prompt)
    {
        String result = chatModelDeepSeek.chat(prompt);

        System.out.println("通过langchain4j调用模型返回结果：\n"+result);

        return result;
    }
}

2.2.3. 测试

访问http://localhost:9002/multimodel/qwen

image

访问http://localhost:9002/multimodel/deepseek

image

2.3.  langchain4j与spring boot的整合

2.3.1. 说明

可以使用原生或者与springboot整合的依赖。

image

两者都有高阶和低阶的版本。

高阶与低阶的对比

特性低阶 API (Low-Level)高阶 API (High-Level)
抽象层级低，接近基础组件高，封装了完整模式
代码量多，需要手动编排少，非常简洁
控制力强，可控制每个细节弱，依赖框架的约定
灵活性高，易于深度定制低，适合标准场景
上手难度较高，需要理解各组件低，非常容易上手
适用场景实现复杂、非标准的逻辑；学习框架原理快速原型开发；构建符合常见模式的应用（如聊天、RAG）

如何选择

  • 优先使用高阶API (**AiServices**)：对于绝大多数常见的应用场景，如构建聊天机器人、文档问答系统等，AiServices都是首选。它能让你以最快的速度和最少的代码量完成开发。

  • 当高阶API无法满足需求时，回退到低阶API：如果你需要实现非常精细的提示词工程、自定义的对话记忆策略，或者要将多个模型和工具以一种非标准的方式组合在一起时，就应该使用低阶API来手动编排整个流程。

2.3.2. 准备

依赖

        <dependency>
            <groupId>dev.langchain4j</groupId>
            <artifactId>langchain4j-open-ai-spring-boot-starter</artifactId>
        </dependency>
        <!--2 LangChain4j 整合boot高阶支持-->
        <dependency>
            <groupId>dev.langchain4j</groupId>
            <artifactId>langchain4j-spring-boot-starter</artifactId>
        </dependency>

配置文件

server.port=9003

spring.application.name=langchain4j-12boot-integration

# https://docs.langchain4j.dev/tutorials/spring-boot-integration
langchain4j.open-ai.chat-model.api-key=sk-cf4de7068032411b9de71adde80e3b12
langchain4j.open-ai.chat-model.model-name=qwen-plus
langchain4j.open-ai.chat-model.base-url=https://dashscope.aliyuncs.com/compatible-mode/v1

2.3.3. 代码

2.3.3.1. 低阶

controller

import dev.langchain4j.model.chat.ChatModel;
import jakarta.annotation.Resource;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RequestParam;
import org.springframework.web.bind.annotation.RestController;

@RestController
public class PopularIntegrationController
{
    @Resource
    private ChatModel chatModel;

    // http://localhost:9003/lc4j/boot/chat
    @GetMapping(value = "/lc4j/boot/chat")
    public String chat(@RequestParam(value = "prompt", defaultValue = "你是谁") String prompt)
    {
        return chatModel.chat(prompt);
    }
}
2.3.3.2. 高阶

controller

import com.atguigu.study.service.ChatAssistant;
import jakarta.annotation.Resource;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RequestParam;
import org.springframework.web.bind.annotation.RestController;

/**
 * @auther zzyybs@126.com
 * @Date 2025-06-18 15:43
 * @Description: TODO
 */
@RestController
public class DeclarativeAIServiceController
{
    @Resource
    private ChatAssistant chatAssistantQwen;

    // http://localhost:9003/lc4j/boot/declarative
    @GetMapping(value = "/lc4j/boot/declarative")
    public String declarative(@RequestParam(value = "prompt", defaultValue = "你是谁") String prompt)
    {
        return chatAssistantQwen.chat(prompt);
    }
}

service接口

import dev.langchain4j.service.spring.AiService;
import dev.langchain4j.service.spring.AiServiceWiringMode;
import org.springframework.stereotype.Service;

@AiService
public interface ChatAssistant
{
    String chat(String prompt);
}

2.3.4. 测试

访问http://localhost:9003/lc4j/boot/chat测试低阶

image

访问http://localhost:9003/lc4j/boot/declarative测试高阶

image

3. 高阶低阶API使用

低阶 API (Low-Level APIs)

低阶API指的是LangChain4j中那些基础的、独立的构建模块。使用它们时，你需要手动将这些模块“连接”或“编排”在一起，以实现一个完整的功能。

特点：

  • 完全的控制力：你可以精确控制流程的每一步，例如如何构建提示词、何时与模型交互、如何管理记忆等。

  • 最高的灵活性：可以实现非常规或高度定制化的逻辑。

  • 代码更冗长：需要编写更多的“胶水代码”来连接各个组件。

  • 更清晰的流程：对于学习其内部工作原理非常有帮助，因为每一步都是显式调用的。

高阶 API (High-Level APIs)

高阶API是构建在低阶组件之上的便捷抽象。它们将常见的应用模式（如对话、RAG）封装起来，提供一个极其简洁的接口。

特点：

  • 极其简洁：用最少的代码实现复杂功能，极大提升开发效率。

  • 约定优于配置：框架帮你处理了很多细节，你只需要关注业务逻辑。

  • 易于上手：对新手非常友好，可以快速构建出原型。

  • 灵活性较低：如果你的需求偏离了它预设的模式，定制起来会比较困难。

3.1. 低阶API使用

3.1.1. 准备

配置类

import com.atguigu.study.service.ChatAssistant;
import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.openai.OpenAiChatModel;
import dev.langchain4j.service.AiServices;
import org.springframework.beans.factory.annotation.Qualifier;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-27 22:04
 * @Description: 知识出处 https://docs.langchain4j.dev/get-started
 */
@Configuration
public class LLMConfig
{
    @Bean(name = "qwen")
    public ChatModel chatModelQwen()
    {
        return OpenAiChatModel.builder()
                    .apiKey("sk-cf4de7068032411b9de71adde80e3b12")
                    .modelName("qwen-plus")
                    .baseUrl("https://dashscope.aliyuncs.com/compatible-mode/v1")
                .build();
    }

    /**
    * @Description: 知识出处，https://api-docs.deepseek.com/zh-cn/
    * @Auther: zzyybs@126.com
    */
    @Bean(name = "deepseek")
    public ChatModel chatModelDeepSeek()
    {
        return
                OpenAiChatModel.builder()
                        .apiKey("sk-fbe1701f86464a11bfa59b8dfbcf7f69")
                        .modelName("deepseek-chat")
                        //.modelName("deepseek-reasoner")
                        .baseUrl("https://api.deepseek.com/v1")
                .build();
    }

}

3.1.2. 代码

使用DeepSeek实现模型问答token用量计算

controller

import dev.langchain4j.data.message.ChatMessage;
import dev.langchain4j.data.message.ChatMessageType;
import dev.langchain4j.data.message.UserMessage;
import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.chat.request.ChatRequest;
import dev.langchain4j.model.chat.response.ChatResponse;
import dev.langchain4j.model.output.TokenUsage;
import jakarta.annotation.Resource;
import lombok.extern.slf4j.Slf4j;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RequestParam;
import org.springframework.web.bind.annotation.RestController;

@RestController
@Slf4j
public class LowApiController
{
    @Resource(name = "deepseek")
    private ChatModel chatModelDeepSeek;

    @GetMapping(value = "/lowapi/api02")
    public String api02(@RequestParam(value = "prompt", defaultValue = "你是谁") String prompt)
    {
        ChatResponse chatResponse = chatModelDeepSeek.chat(UserMessage.from(prompt));

        String result = chatResponse.aiMessage().text();
        System.out.println("通过调用大模型返回结果："+result);

        // Token 用量计算的底层api
        TokenUsage tokenUsage = chatResponse.tokenUsage();

        System.out.println("本次调用消耗的token："+tokenUsage);

        result = result +"\t\n"+tokenUsage;

        return result;
    }
}

3.1.3. 测试

image

3.2. 高阶API使用

使用千问模型实现高阶API调用

3.3. 代码

配置类

import com.atguigu.study.service.ChatAssistant;
import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.openai.OpenAiChatModel;
import dev.langchain4j.service.AiServices;
import org.springframework.beans.factory.annotation.Qualifier;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-27 22:04
 * @Description: 知识出处 https://docs.langchain4j.dev/get-started
 */
@Configuration
public class LLMConfig
{
    @Bean(name = "qwen")
    public ChatModel chatModelQwen()
    {
        return OpenAiChatModel.builder()
                    .apiKey("sk-cf4de7068032411b9de71adde80e3b12")
                    .modelName("qwen-plus")
                    .baseUrl("https://dashscope.aliyuncs.com/compatible-mode/v1")
                .build();
    }

    /**
    * @Description: 知识出处，https://api-docs.deepseek.com/zh-cn/
    * @Auther: zzyybs@126.com
    */
    @Bean(name = "deepseek")
    public ChatModel chatModelDeepSeek()
    {
        return
                OpenAiChatModel.builder()
                        .apiKey("sk-fbe1701f86464a11bfa59b8dfbcf7f69")
                        .modelName("deepseek-chat")
                        //.modelName("deepseek-reasoner")
                        .baseUrl("https://api.deepseek.com/v1")
                .build();
    }



    // High-Api https://docs.langchain4j.dev/tutorials/ai-services#simplest-ai-service
    @Bean
    public ChatAssistant chatAssistant(@Qualifier("qwen") ChatModel chatModelQwen)
    {
        return AiServices.create(ChatAssistant.class, chatModelQwen);
    }
}

controller

import com.atguigu.study.service.ChatAssistant;
import dev.langchain4j.model.chat.ChatModel;
import jakarta.annotation.Resource;
import lombok.extern.slf4j.Slf4j;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RequestParam;
import org.springframework.web.bind.annotation.RestController;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-28 17:32
 * @Description: TODO
 */
@RestController
@Slf4j
public class HighApiController
{
    @Resource
    private ChatAssistant chatAssistant;

    @GetMapping(value = "/highapi/api01")
    public String highApi(@RequestParam(value = "prompt", defaultValue = "你是谁") String prompt)
    {
        return chatAssistant.chat(prompt);
    }
}

service

/**
 * @Description: 我们知道，按照Java开发一般习惯，有接口就要有实现类
 * 比如接口ChatAssistant，就会有实现类ChatAssistantImpl
 * 现在用高阶api-AIServics不用你自己写impl实现类，交给langchain4j给你搞定
 *
 * 本次配置用的是langchain4j原生整合，没有引入sprinboot，不需要接口头上配置@AiService注解标签
 */
public interface ChatAssistant
{
    String chat(String prompt);
}

3.4. 测试

访问：http://localhost:9004/highapi/api01?prompt=1+1=等于几

image

4. 模型参数配置

通过配置一些参数可以自定义模型模型实现一些高级功能。

4.1. 日志

4.1.1. 准备

依赖

<dependency>
  <groupId>ch.qos.logback</groupId>
  <artifactId>logback-classic</artifactId>
  <version>1.5.8</version>
</dependency>

如果使用spring boot的langchain4j包，需要在配置文件中加入

logging.level.dev.langchain4j = DEBUG

4.1.2. 代码

controller


import dev.langchain4j.data.message.UserMessage;
import dev.langchain4j.model.chat.ChatModel;
import jakarta.annotation.Resource;
import lombok.extern.slf4j.Slf4j;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RequestParam;
import org.springframework.web.bind.annotation.RestController;

import java.util.concurrent.TimeUnit;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-28 18:33
 * @Description: 知识出处，https://docs.langchain4j.dev/tutorials/model-parameters/
 */
@RestController
@Slf4j
public class ModelParameterController
{
    @Resource
    private ChatModel chatModelQwen;

    // http://localhost:9005/modelparam/config
    @GetMapping(value = "/modelparam/config")
    public String config(@RequestParam(value = "prompt", defaultValue = "你是谁") String prompt)
    {
        String result = chatModelQwen.chat(prompt);

        System.out.println("通过langchain4j调用模型返回结果："+result);

        return result;
    }
}

config

import com.atguigu.study.listener.TestChatModelListener;
import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.openai.OpenAiChatModel;
import dev.langchain4j.service.AiServices;
import org.springframework.beans.factory.annotation.Qualifier;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;

import java.time.Duration;
import java.util.List;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-27 22:04
 * @Description: 知识出处 https://docs.langchain4j.dev/tutorials/model-parameters/
 */
@Configuration
public class LLMConfig
{
    @Bean(name = "qwen")
    public ChatModel chatModelQwen()
    {
        return OpenAiChatModel.builder()
                    .apiKey("sk-cf4de7068032411b9de71adde80e3b12")
                    .modelName("qwen-plus")
                    .baseUrl("https://dashscope.aliyuncs.com/compatible-mode/v1")
                .logRequests(true)
                .logResponses(true)
                .build();
    }
}

4.1.3. 测试

访问http://localhost:9005/modelparam/config后台会打印日志。

image

4.2. 监控

4.2.1. 说明

类似于spring中的环绕通知。继承langchain4j提供的listener，实现ChatModeListener来监听以下事件：

  • 向LLM提出的请求

  • LLM的回应

  • 错误

ChatModeListener

import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.chat.request.ChatRequest;
import dev.langchain4j.model.chat.response.ChatResponse;

/**
 * A {@link ChatModel} listener that listens for requests, responses and errors.
 */
public interface ChatModelListener {

    /**
     * This method is called before the request is sent to the model.
     *
     * @param requestContext The request context. It contains the {@link ChatRequest} and attributes.
     *                       The attributes can be used to pass data between methods of this listener
     *                       or between multiple listeners.
     */
    default void onRequest(ChatModelRequestContext requestContext) {

    }

    /**
     * This method is called after the response is received from the model.
     *
     * @param responseContext The response context.
     *                        It contains {@link ChatResponse}, corresponding {@link ChatRequest} and attributes.
     *                        The attributes can be used to pass data between methods of this listener
     *                        or between multiple listeners.
     */
    default void onResponse(ChatModelResponseContext responseContext) {

    }

    /**
     * This method is called when an error occurs during interaction with the model.
     *
     * @param errorContext The error context.
     *                     It contains the error, corresponding {@link ChatRequest},
     *                     partial {@link ChatResponse} (if available) and attributes.
     *                     The attributes can be used to pass data between methods of this listener
     *                     or between multiple listeners.
     */
    default void onError(ChatModelErrorContext errorContext) {

    }
}

4.2.2. 准备

创建TestChatModelListener实现ChatModeListener接口。访问LLM前生成一个随机的字符串存储在ChatModelRequestContext对象的attributes中

imageTestChatModelListener

import cn.hutool.core.util.IdUtil;
import dev.langchain4j.model.chat.listener.ChatModelErrorContext;
import dev.langchain4j.model.chat.listener.ChatModelListener;
import dev.langchain4j.model.chat.listener.ChatModelRequestContext;
import dev.langchain4j.model.chat.listener.ChatModelResponseContext;
import lombok.extern.slf4j.Slf4j;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-28 18:53
 * @Description: 知识出处，https://docs.langchain4j.dev/tutorials/spring-boot-integration#observability
 */
@Slf4j
public class TestChatModelListener implements ChatModelListener
{
    @Override
    public void onRequest(ChatModelRequestContext requestContext)
    {

        // onRequest配置的k:v键值对，在onResponse阶段可以获得，上下文传递参数好用
        String uuidValue = IdUtil.simpleUUID();
        requestContext.attributes().put("TraceID",uuidValue);
        log.info("请求参数requestContext:{}", requestContext+"\t"+uuidValue);
    }

    @Override
    public void onResponse(ChatModelResponseContext responseContext)
    {
        Object object = responseContext.attributes().get("TraceID");

        log.info("返回结果responseContext:{}", object);
    }

    @Override
    public void onError(ChatModelErrorContext errorContext)
    {
        log.error("请求异常ChatModelErrorContext:{}", errorContext);
    }
}

4.2.3. 代码

config

import com.atguigu.study.listener.TestChatModelListener;
import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.openai.OpenAiChatModel;
import dev.langchain4j.service.AiServices;
import org.springframework.beans.factory.annotation.Qualifier;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;

import java.time.Duration;
import java.util.List;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-27 22:04
 * @Description: 知识出处 https://docs.langchain4j.dev/tutorials/model-parameters/
 */
@Configuration
public class LLMConfig
{
    @Bean(name = "qwen")
    public ChatModel chatModelQwen()
    {
        return OpenAiChatModel.builder()
                    .apiKey("sk-cf4de7068032411b9de71adde80e3b12")
                    .modelName("qwen-plus")
                    .baseUrl("https://dashscope.aliyuncs.com/compatible-mode/v1")
                .logRequests(true)
                .logResponses(true)
                .listeners(List.of(new TestChatModelListener()))
                .build();
    }
}

4.2.4. 测试

访问http://localhost:9005/modelparam/config

image

4.3. 重试机制

4.3.1. 说明

当使用langchain4j访问大模型时，可能会出现访问中断的情况，设置重试参数，可以在访问大模型失败后自动重试。

4.3.2. 代码

import com.atguigu.study.listener.TestChatModelListener;
import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.openai.OpenAiChatModel;
import dev.langchain4j.service.AiServices;
import org.springframework.beans.factory.annotation.Qualifier;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;

import java.time.Duration;
import java.util.List;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-27 22:04
 * @Description: 知识出处 https://docs.langchain4j.dev/tutorials/model-parameters/
 */
@Configuration
public class LLMConfig
{
    @Bean(name = "qwen")
    public ChatModel chatModelQwen()
    {
        return OpenAiChatModel.builder()
                    .apiKey("sk-cf4de7068032411b9de71adde80e3b12")
                    .modelName("qwen-plus")
                    .baseUrl("https://dashscope.aliyuncs.com/compatible-mode/v1")
                .logRequests(true)
                .logResponses(true)
                .listeners(List.of(new TestChatModelListener()))
                .maxRetries(2) //发送两次请求
                .build();
    }
}

4.3.3. 测试

断开网络连接，访问http://localhost:9005/modelparam/config

控制台输出

2025-07-17T14:26:34.068+08:00  INFO 32588 --- [langchain4j-05model-parameters] [nio-9005-exec-4] c.a.s.listener.TestChatModelListener     : 请求参数requestContext:dev.langchain4j.model.chat.listener.ChatModelRequestContext@19ea835b	3b16258519f64cf3a9b462f0517090ea
2025-07-17T14:26:34.068+08:00  INFO 32588 --- [langchain4j-05model-parameters] [nio-9005-exec-4] d.l.http.client.log.LoggingHttpClient    : HTTP request:
- method: POST
- url: https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions
- headers: [Authorization: Beare...12], [User-Agent: langchain4j-openai], [Content-Type: application/json]
- body: {
  "model" : "qwen-plus",
  "messages" : [ {
    "role" : "user",
    "content" : "你是谁"
  } ],
  "stream" : false
}

2025-07-17T14:26:53.256+08:00  WARN 32588 --- [langchain4j-05model-parameters] [nio-9005-exec-4] dev.langchain4j.internal.RetryUtils      : A retriable exception occurred. Remaining retries: 2 of 2

java.lang.RuntimeException: java.io.IOException: Connection reset
	at dev.langchain4j.http.client.jdk.JdkHttpClient.execute(JdkHttpClient.java:60) ~[langchain4j-http-client-jdk-1.0.1.jar:na]
	at dev.langchain4j.http.client.log.LoggingHttpClient.execute(LoggingHttpClient.java:39) ~[langchain4j-http-client-1.0.1.jar:na]
	at dev.langchain4j.model.openai.internal.SyncRequestExecutor.execute(SyncRequestExecutor.java:20) ~[langchain4j-open-ai-1.0.1.jar:na]
	at dev.langchain4j.model.openai.internal.RequestExecutor.execute(RequestExecutor.java:39) ~[langchain4j-open-ai-1.0.1.jar:na]
	at dev.langchain4j.model.openai.OpenAiChatModel.lambda$doChat$0(OpenAiChatModel.java:143) ~[langchain4j-open-ai-1.0.1.jar:na]
	at dev.langchain4j.internal.ExceptionMapper.withExceptionMapper(ExceptionMapper.java:29) ~[langchain4j-core-1.0.1.jar:na]
	at dev.langchain4j.internal.RetryUtils.lambda$withRetryMappingExceptions$2(RetryUtils.java:324) ~[langchain4j-core-1.0.1.jar:na]
	at dev.langchain4j.internal.RetryUtils$RetryPolicy.withRetry(RetryUtils.java:211) ~[langchain4j-core-1.0.1.jar:na]
	at dev.langchain4j.internal.RetryUtils.withRetry(RetryUtils.java:264) ~[langchain4j-core-1.0.1.jar:na]
	at dev.langchain4j.internal.RetryUtils.withRetryMappingExceptions(RetryUtils.java:324) ~[langchain4j-core-1.0.1.jar:na]
	at dev.langchain4j.internal.RetryUtils.withRetryMappingExceptions(RetryUtils.java:308) ~[langchain4j-core-1.0.1.jar:na]
	at dev.langchain4j.model.openai.OpenAiChatModel.doChat(OpenAiChatModel.java:142) ~[langchain4j-open-ai-1.0.1.jar:na]
	at dev.langchain4j.model.chat.ChatModel.chat(ChatModel.java:46) ~[langchain4j-core-1.0.1.jar:na]
	at dev.langchain4j.model.chat.ChatModel.chat(ChatModel.java:77) ~[langchain4j-core-1.0.1.jar:na]
	at com.atguigu.study.controller.ModelParameterController.config(ModelParameterController.java:29) ~[classes/:na]
	at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method) ~[na:na]
	at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:77) ~[na:na]
	at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) ~[na:na]
	at java.base/java.lang.reflect.Method.invoke(Method.java:568) ~[na:na]
	at org.springframework.web.method.support.InvocableHandlerMethod.doInvoke(InvocableHandlerMethod.java:258) ~[spring-web-6.2.7.jar:6.2.7]
	at org.springframework.web.method.support.InvocableHandlerMethod.invokeForRequest(InvocableHandlerMethod.java:191) ~[spring-web-6.2.7.jar:6.2.7]
	at org.springframework.web.servlet.mvc.method.annotation.ServletInvocableHandlerMethod.invokeAndHandle(ServletInvocableHandlerMethod.java:118) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at org.springframework.web.servlet.mvc.method.annotation.RequestMappingHandlerAdapter.invokeHandlerMethod(RequestMappingHandlerAdapter.java:986) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at org.springframework.web.servlet.mvc.method.annotation.RequestMappingHandlerAdapter.handleInternal(RequestMappingHandlerAdapter.java:891) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at org.springframework.web.servlet.mvc.method.AbstractHandlerMethodAdapter.handle(AbstractHandlerMethodAdapter.java:87) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at org.springframework.web.servlet.DispatcherServlet.doDispatch(DispatcherServlet.java:1089) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at org.springframework.web.servlet.DispatcherServlet.doService(DispatcherServlet.java:979) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at org.springframework.web.servlet.FrameworkServlet.processRequest(FrameworkServlet.java:1014) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at org.springframework.web.servlet.FrameworkServlet.doGet(FrameworkServlet.java:903) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at jakarta.servlet.http.HttpServlet.service(HttpServlet.java:564) ~[tomcat-embed-core-10.1.41.jar:6.0]
	at org.springframework.web.servlet.FrameworkServlet.service(FrameworkServlet.java:885) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at jakarta.servlet.http.HttpServlet.service(HttpServlet.java:658) ~[tomcat-embed-core-10.1.41.jar:6.0]
	at org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:195) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:140) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.tomcat.websocket.server.WsFilter.doFilter(WsFilter.java:51) ~[tomcat-embed-websocket-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:164) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:140) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.springframework.web.filter.RequestContextFilter.doFilterInternal(RequestContextFilter.java:100) ~[spring-web-6.2.7.jar:6.2.7]
	at org.springframework.web.filter.OncePerRequestFilter.doFilter(OncePerRequestFilter.java:116) ~[spring-web-6.2.7.jar:6.2.7]
	at org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:164) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:140) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.springframework.web.filter.FormContentFilter.doFilterInternal(FormContentFilter.java:93) ~[spring-web-6.2.7.jar:6.2.7]
	at org.springframework.web.filter.OncePerRequestFilter.doFilter(OncePerRequestFilter.java:116) ~[spring-web-6.2.7.jar:6.2.7]
	at org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:164) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:140) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.springframework.web.filter.CharacterEncodingFilter.doFilterInternal(CharacterEncodingFilter.java:201) ~[spring-web-6.2.7.jar:6.2.7]
	at org.springframework.web.filter.OncePerRequestFilter.doFilter(OncePerRequestFilter.java:116) ~[spring-web-6.2.7.jar:6.2.7]
	at org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:164) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:140) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.StandardWrapperValve.invoke(StandardWrapperValve.java:167) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.StandardContextValve.invoke(StandardContextValve.java:90) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.authenticator.AuthenticatorBase.invoke(AuthenticatorBase.java:483) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.StandardHostValve.invoke(StandardHostValve.java:116) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.valves.ErrorReportValve.invoke(ErrorReportValve.java:93) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.StandardEngineValve.invoke(StandardEngineValve.java:74) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.connector.CoyoteAdapter.service(CoyoteAdapter.java:344) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.coyote.http11.Http11Processor.service(Http11Processor.java:398) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.coyote.AbstractProcessorLight.process(AbstractProcessorLight.java:63) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.coyote.AbstractProtocol$ConnectionHandler.process(AbstractProtocol.java:903) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.tomcat.util.net.NioEndpoint$SocketProcessor.doRun(NioEndpoint.java:1740) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.tomcat.util.net.SocketProcessorBase.run(SocketProcessorBase.java:52) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.tomcat.util.threads.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1189) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.tomcat.util.threads.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:658) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.tomcat.util.threads.TaskThread$WrappingRunnable.run(TaskThread.java:63) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at java.base/java.lang.Thread.run(Thread.java:833) ~[na:na]
Caused by: java.io.IOException: Connection reset
	at java.net.http/jdk.internal.net.http.HttpClientImpl.send(HttpClientImpl.java:586) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.HttpClientFacade.send(HttpClientFacade.java:123) ~[java.net.http:na]
	at dev.langchain4j.http.client.jdk.JdkHttpClient.execute(JdkHttpClient.java:50) ~[langchain4j-http-client-jdk-1.0.1.jar:na]
	... 64 common frames omitted
Caused by: java.net.SocketException: Connection reset
	at java.base/sun.nio.ch.SocketChannelImpl.throwConnectionReset(SocketChannelImpl.java:394) ~[na:na]
	at java.base/sun.nio.ch.SocketChannelImpl.read(SocketChannelImpl.java:426) ~[na:na]
	at java.net.http/jdk.internal.net.http.SocketTube.readAvailable(SocketTube.java:1170) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.SocketTube$InternalReadPublisher$InternalReadSubscription.read(SocketTube.java:833) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.SocketTube$SocketFlowTask.run(SocketTube.java:181) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.common.SequentialScheduler$SchedulableTask.run(SequentialScheduler.java:230) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.common.SequentialScheduler.runOrSchedule(SequentialScheduler.java:303) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.common.SequentialScheduler.runOrSchedule(SequentialScheduler.java:256) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.SocketTube$InternalReadPublisher$InternalReadSubscription.signalReadable(SocketTube.java:774) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.SocketTube$InternalReadPublisher$ReadEvent.signalEvent(SocketTube.java:957) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.SocketTube$SocketFlowEvent.handle(SocketTube.java:253) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.HttpClientImpl$SelectorManager.handleEvent(HttpClientImpl.java:979) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.HttpClientImpl$SelectorManager.lambda$run$3(HttpClientImpl.java:934) ~[java.net.http:na]
	at java.base/java.util.ArrayList.forEach(ArrayList.java:1511) ~[na:na]
	at java.net.http/jdk.internal.net.http.HttpClientImpl$SelectorManager.run(HttpClientImpl.java:934) ~[java.net.http:na]

2025-07-17T14:26:53.849+08:00  INFO 32588 --- [langchain4j-05model-parameters] [nio-9005-exec-4] d.l.http.client.log.LoggingHttpClient    : HTTP request:
- method: POST
- url: https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions
- headers: [Authorization: Beare...12], [User-Agent: langchain4j-openai], [Content-Type: application/json]
- body: {
  "model" : "qwen-plus",
  "messages" : [ {
    "role" : "user",
    "content" : "你是谁"
  } ],
  "stream" : false
}

2025-07-17T14:26:53.854+08:00 ERROR 32588 --- [langchain4j-05model-parameters] [nio-9005-exec-4] c.a.s.listener.TestChatModelListener     : 请求异常ChatModelErrorContext:dev.langchain4j.model.chat.listener.ChatModelErrorContext@35277b72
2025-07-17T14:26:53.859+08:00 ERROR 32588 --- [langchain4j-05model-parameters] [nio-9005-exec-4] o.a.c.c.C.[.[.[/].[dispatcherServlet]    : Servlet.service() for servlet [dispatcherServlet] in context with path [] threw exception [Request processing failed: dev.langchain4j.exception.UnresolvedModelServerException] with root cause

java.nio.channels.UnresolvedAddressException: null
	at java.base/sun.nio.ch.Net.checkAddress(Net.java:149) ~[na:na]
	at java.base/sun.nio.ch.Net.checkAddress(Net.java:157) ~[na:na]
	at java.base/sun.nio.ch.SocketChannelImpl.checkRemote(SocketChannelImpl.java:816) ~[na:na]
	at java.base/sun.nio.ch.SocketChannelImpl.connect(SocketChannelImpl.java:839) ~[na:na]
	at java.net.http/jdk.internal.net.http.PlainHttpConnection.lambda$connectAsync$0(PlainHttpConnection.java:183) ~[java.net.http:na]
	at java.base/java.security.AccessController.doPrivileged(AccessController.java:569) ~[na:na]
	at java.net.http/jdk.internal.net.http.PlainHttpConnection.connectAsync(PlainHttpConnection.java:185) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.AsyncSSLConnection.connectAsync(AsyncSSLConnection.java:56) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.Http2Connection.createAsync(Http2Connection.java:378) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.Http2ClientImpl.getConnectionFor(Http2ClientImpl.java:128) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.ExchangeImpl.get(ExchangeImpl.java:93) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.Exchange.establishExchange(Exchange.java:343) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.Exchange.responseAsyncImpl0(Exchange.java:475) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.Exchange.responseAsyncImpl(Exchange.java:380) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.Exchange.responseAsync(Exchange.java:372) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.MultiExchange.responseAsyncImpl(MultiExchange.java:408) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.MultiExchange.lambda$responseAsyncImpl$7(MultiExchange.java:449) ~[java.net.http:na]
	at java.base/java.util.concurrent.CompletableFuture.uniHandle(CompletableFuture.java:934) ~[na:na]
	at java.base/java.util.concurrent.CompletableFuture.uniHandleStage(CompletableFuture.java:950) ~[na:na]
	at java.base/java.util.concurrent.CompletableFuture.handle(CompletableFuture.java:2340) ~[na:na]
	at java.net.http/jdk.internal.net.http.MultiExchange.responseAsyncImpl(MultiExchange.java:439) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.MultiExchange.lambda$responseAsync0$2(MultiExchange.java:341) ~[java.net.http:na]
	at java.base/java.util.concurrent.CompletableFuture$UniCompose.tryFire(CompletableFuture.java:1150) ~[na:na]
	at java.base/java.util.concurrent.CompletableFuture.postComplete(CompletableFuture.java:510) ~[na:na]
	at java.base/java.util.concurrent.CompletableFuture$AsyncSupply.run(CompletableFuture.java:1773) ~[na:na]
	at java.net.http/jdk.internal.net.http.HttpClientImpl$DelegatingExecutor.execute(HttpClientImpl.java:157) ~[java.net.http:na]
	at java.base/java.util.concurrent.CompletableFuture.completeAsync(CompletableFuture.java:2673) ~[na:na]
	at java.net.http/jdk.internal.net.http.MultiExchange.responseAsync(MultiExchange.java:294) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.HttpClientImpl.sendAsync(HttpClientImpl.java:654) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.HttpClientImpl.send(HttpClientImpl.java:552) ~[java.net.http:na]
	at java.net.http/jdk.internal.net.http.HttpClientFacade.send(HttpClientFacade.java:123) ~[java.net.http:na]
	at dev.langchain4j.http.client.jdk.JdkHttpClient.execute(JdkHttpClient.java:50) ~[langchain4j-http-client-jdk-1.0.1.jar:na]
	at dev.langchain4j.http.client.log.LoggingHttpClient.execute(LoggingHttpClient.java:39) ~[langchain4j-http-client-1.0.1.jar:na]
	at dev.langchain4j.model.openai.internal.SyncRequestExecutor.execute(SyncRequestExecutor.java:20) ~[langchain4j-open-ai-1.0.1.jar:na]
	at dev.langchain4j.model.openai.internal.RequestExecutor.execute(RequestExecutor.java:39) ~[langchain4j-open-ai-1.0.1.jar:na]
	at dev.langchain4j.model.openai.OpenAiChatModel.lambda$doChat$0(OpenAiChatModel.java:143) ~[langchain4j-open-ai-1.0.1.jar:na]
	at dev.langchain4j.internal.ExceptionMapper.withExceptionMapper(ExceptionMapper.java:29) ~[langchain4j-core-1.0.1.jar:na]
	at dev.langchain4j.internal.RetryUtils.lambda$withRetryMappingExceptions$2(RetryUtils.java:324) ~[langchain4j-core-1.0.1.jar:na]
	at dev.langchain4j.internal.RetryUtils$RetryPolicy.withRetry(RetryUtils.java:211) ~[langchain4j-core-1.0.1.jar:na]
	at dev.langchain4j.internal.RetryUtils.withRetry(RetryUtils.java:264) ~[langchain4j-core-1.0.1.jar:na]
	at dev.langchain4j.internal.RetryUtils.withRetryMappingExceptions(RetryUtils.java:324) ~[langchain4j-core-1.0.1.jar:na]
	at dev.langchain4j.internal.RetryUtils.withRetryMappingExceptions(RetryUtils.java:308) ~[langchain4j-core-1.0.1.jar:na]
	at dev.langchain4j.model.openai.OpenAiChatModel.doChat(OpenAiChatModel.java:142) ~[langchain4j-open-ai-1.0.1.jar:na]
	at dev.langchain4j.model.chat.ChatModel.chat(ChatModel.java:46) ~[langchain4j-core-1.0.1.jar:na]
	at dev.langchain4j.model.chat.ChatModel.chat(ChatModel.java:77) ~[langchain4j-core-1.0.1.jar:na]
	at com.atguigu.study.controller.ModelParameterController.config(ModelParameterController.java:29) ~[classes/:na]
	at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method) ~[na:na]
	at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:77) ~[na:na]
	at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) ~[na:na]
	at java.base/java.lang.reflect.Method.invoke(Method.java:568) ~[na:na]
	at org.springframework.web.method.support.InvocableHandlerMethod.doInvoke(InvocableHandlerMethod.java:258) ~[spring-web-6.2.7.jar:6.2.7]
	at org.springframework.web.method.support.InvocableHandlerMethod.invokeForRequest(InvocableHandlerMethod.java:191) ~[spring-web-6.2.7.jar:6.2.7]
	at org.springframework.web.servlet.mvc.method.annotation.ServletInvocableHandlerMethod.invokeAndHandle(ServletInvocableHandlerMethod.java:118) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at org.springframework.web.servlet.mvc.method.annotation.RequestMappingHandlerAdapter.invokeHandlerMethod(RequestMappingHandlerAdapter.java:986) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at org.springframework.web.servlet.mvc.method.annotation.RequestMappingHandlerAdapter.handleInternal(RequestMappingHandlerAdapter.java:891) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at org.springframework.web.servlet.mvc.method.AbstractHandlerMethodAdapter.handle(AbstractHandlerMethodAdapter.java:87) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at org.springframework.web.servlet.DispatcherServlet.doDispatch(DispatcherServlet.java:1089) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at org.springframework.web.servlet.DispatcherServlet.doService(DispatcherServlet.java:979) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at org.springframework.web.servlet.FrameworkServlet.processRequest(FrameworkServlet.java:1014) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at org.springframework.web.servlet.FrameworkServlet.doGet(FrameworkServlet.java:903) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at jakarta.servlet.http.HttpServlet.service(HttpServlet.java:564) ~[tomcat-embed-core-10.1.41.jar:6.0]
	at org.springframework.web.servlet.FrameworkServlet.service(FrameworkServlet.java:885) ~[spring-webmvc-6.2.7.jar:6.2.7]
	at jakarta.servlet.http.HttpServlet.service(HttpServlet.java:658) ~[tomcat-embed-core-10.1.41.jar:6.0]
	at org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:195) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:140) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.tomcat.websocket.server.WsFilter.doFilter(WsFilter.java:51) ~[tomcat-embed-websocket-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:164) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:140) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.springframework.web.filter.RequestContextFilter.doFilterInternal(RequestContextFilter.java:100) ~[spring-web-6.2.7.jar:6.2.7]
	at org.springframework.web.filter.OncePerRequestFilter.doFilter(OncePerRequestFilter.java:116) ~[spring-web-6.2.7.jar:6.2.7]
	at org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:164) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:140) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.springframework.web.filter.FormContentFilter.doFilterInternal(FormContentFilter.java:93) ~[spring-web-6.2.7.jar:6.2.7]
	at org.springframework.web.filter.OncePerRequestFilter.doFilter(OncePerRequestFilter.java:116) ~[spring-web-6.2.7.jar:6.2.7]
	at org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:164) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:140) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.springframework.web.filter.CharacterEncodingFilter.doFilterInternal(CharacterEncodingFilter.java:201) ~[spring-web-6.2.7.jar:6.2.7]
	at org.springframework.web.filter.OncePerRequestFilter.doFilter(OncePerRequestFilter.java:116) ~[spring-web-6.2.7.jar:6.2.7]
	at org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:164) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:140) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.StandardWrapperValve.invoke(StandardWrapperValve.java:167) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.StandardContextValve.invoke(StandardContextValve.java:90) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.authenticator.AuthenticatorBase.invoke(AuthenticatorBase.java:483) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.StandardHostValve.invoke(StandardHostValve.java:116) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.valves.ErrorReportValve.invoke(ErrorReportValve.java:93) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.core.StandardEngineValve.invoke(StandardEngineValve.java:74) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.catalina.connector.CoyoteAdapter.service(CoyoteAdapter.java:344) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.coyote.http11.Http11Processor.service(Http11Processor.java:398) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.coyote.AbstractProcessorLight.process(AbstractProcessorLight.java:63) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.coyote.AbstractProtocol$ConnectionHandler.process(AbstractProtocol.java:903) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.tomcat.util.net.NioEndpoint$SocketProcessor.doRun(NioEndpoint.java:1740) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.tomcat.util.net.SocketProcessorBase.run(SocketProcessorBase.java:52) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.tomcat.util.threads.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1189) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.tomcat.util.threads.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:658) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at org.apache.tomcat.util.threads.TaskThread$WrappingRunnable.run(TaskThread.java:63) ~[tomcat-embed-core-10.1.41.jar:10.1.41]
	at java.base/java.lang.Thread.run(Thread.java:833) ~[na:na]

4.4. 超时机制

4.4.1. 说明

向大模型发送请求时，如在指定时间内没有收到响应，该请求将被中断并报request timed out

4.4.2. 代码

config

import com.atguigu.study.listener.TestChatModelListener;
import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.openai.OpenAiChatModel;
import dev.langchain4j.service.AiServices;
import org.springframework.beans.factory.annotation.Qualifier;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;

import java.time.Duration;
import java.util.List;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-27 22:04
 * @Description: 知识出处 https://docs.langchain4j.dev/tutorials/model-parameters/
 */
@Configuration
public class LLMConfig
{
    @Bean(name = "qwen")
    public ChatModel chatModelQwen()
    {
        return OpenAiChatModel.builder()
                    .apiKey("sk-cf4de7068032411b9de71adde80e3b12")
                    .modelName("qwen-plus")
                    .baseUrl("https://dashscope.aliyuncs.com/compatible-mode/v1")
                .logRequests(true)
                .logResponses(true)
                .listeners(List.of(new TestChatModelListener()))
                .maxRetries(2)
                .timeout(Duration.ofSeconds(2))//向大模型发送请求时，如在指定时间内没有收到响应，该请求将被中断并报request timed out
                .build();
    }
}

4.4.3. 测试

访问http://localhost:9005/modelparam/config?prompt=介绍一下Java

已知这个请求的响应时间会超过两秒。所以控制台会输出request timed out

image

5. 多模态视觉理解

langchain4j不仅可以支持理解文字，还可以传递解析后的图片、视频、音频、pdf等。当然，在配置大模型时，你需要确保大模型本身能支持解析这些文件。

5.1. 图片理解

按照langchain官网的知道，将图片通过base64转码为字符串

image

5.1.1. 准备

由于我们之前使用的deepseek-chat和qwen-plus都不支持图片理解，所以我们需要更换一个模型。langchain官网docs.langchain4j.dev/integration…给出了一些支持图片理解的大模型可以作为参考，我们使用通义千问的qwen-vl-max模型。

image

qwen-vl-max的使用方式help.aliyun.com/zh/model-st…

image

需要在配置类LLMConfig中修改modelName和baseUrl。api-key保持不变。

图片理解我们需要实现准备一张图片让大模型去分析。我们使用这张图片，将其放在src/main/resources/static/images目录下，并提问大模型：从下面图片中获取来源网站名称，分析股价走势和5月30号股价

image

5.1.2. 代码

LLMConfig

import dev.langchain4j.community.model.dashscope.WanxImageModel;
import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.openai.OpenAiChatModel;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-30 11:24
 * @Description: 知识出处
 * https://docs.langchain4j.dev/tutorials/chat-and-language-models/#image-content
 */
@Configuration
public class LLMConfig
{
    @Bean
    public ChatModel ImageModel() {
        return OpenAiChatModel.builder()
                .apiKey("sk-cf4de7068032411b9de71adde80e3b12")
                //qwen-vl-max 是一个多模态大模型，支持图片和文本的结合输入，适用于视觉-语言任务。
                .modelName("qwen-vl-max")
                .baseUrl("https://dashscope.aliyuncs.com/compatible-mode/v1")
                .build();
    }
}

controller

根据langchain官网的使用方法，我们编写业务类

import dev.langchain4j.data.message.AiMessage;
import dev.langchain4j.data.message.ImageContent;
import dev.langchain4j.data.message.TextContent;
import dev.langchain4j.data.message.UserMessage;
import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.chat.response.ChatResponse;
import dev.langchain4j.model.output.Response;
import lombok.extern.slf4j.Slf4j;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.beans.factory.annotation.Value;
import org.springframework.core.io.Resource;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RestController;

import java.io.IOException;
import java.util.Base64;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-30 11:26
 * @Description: https://docs.langchain4j.dev/tutorials/chat-and-language-models/#multimodality
 */
@RestController
@Slf4j
public class ImageModelController
{
    @Autowired
    private ChatModel chatModel;

    @Value("classpath:static/images/mi.jpg")
    private Resource resource;//import org.springframework.core.io.Resource;

    /**
    * @Description: 通过Base64编码将图片转化为字符串
     *              结合ImageContent和TextContent形成UserMessage一起发送到模型进行处理。
    * @Auther: zzyybs@126.com
     *
     *测试地址：http://localhost:9006/image/call
    */
    @GetMapping(value = "/image/call")
    public String readImageContent() throws IOException
    {
        String result = null;

        //第一步，图片转码：通过Base64编码将图片转化为字符串
        byte[] byteArray = resource.getContentAsByteArray();
        String base64Data = Base64.getEncoder().encodeToString(byteArray);

        //第二步，提示词指定：结合ImageContent和TextContent一起发送到模型进行处理。
        UserMessage userMessage = UserMessage.from(
                TextContent.from("从下面图片中获取来源网站名称，分析股价走势和5月30号股价"),
                ImageContent.from(base64Data, "image/jpg")
        );
        //第三步，API调用：使用OpenAiChatModel来构建请求，并通过chat()方法调用模型。
        //请求内容包括文本提示和图片，模型会根据输入返回分析结果。
        ChatResponse chatResponse = chatModel.chat(userMessage);

        //第四步，解析与输出：从ChatResponse中获取AI大模型的回复，打印出处理后的结果。
        result = chatResponse.aiMessage().text();

        //后台打印
        System.out.println(result);

        //返回前台
        return result;
    }
}

5.1.3. 测试

访问http://localhost:9006/image/call

得到qwen-vl-max大模型的回答

image

5.2. 文生图

通过文字生成一张图片

5.2.1. 准备

使用阿里巴巴通义万象进行图像理解，其支持视觉-语言的多模态任务。

docs.langchain4j.dev/integration…

image

文生图需要用到第三方的api，所以需要引入qwen的依赖

image

5.2.2. 代码

父pom增加依赖

<dependency>
  <groupId>dev.langchain4j</groupId>
  <artifactId>langchain4j-community-bom</artifactId>
  <version>${langchain4j-community.version}</version>
  <type>pom</type>
  <scope>import</scope>
</dependency>

子pom增加依赖

<!--DashScope (Qwen)接入阿里云百炼平台
https://docs.langchain4j.dev/integrations/language-models/dashscope
-->
<dependency>
  <groupId>dev.langchain4j</groupId>
  <artifactId>langchain4j-community-dashscope-spring-boot-starter</artifactId>
</dependency>

config

import dev.langchain4j.community.model.dashscope.WanxImageModel;
import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.openai.OpenAiChatModel;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-30 11:24
 * @Description: 知识出处
 * https://docs.langchain4j.dev/tutorials/chat-and-language-models/#image-content
 */
@Configuration
public class LLMConfig
{
    @Bean
    public ChatModel ImageModel() {
        return OpenAiChatModel.builder()
        .apiKey("sk-cf4de7068032411b9de71adde80e3b12")
        //qwen-vl-max 是一个多模态大模型，支持图片和文本的结合输入，适用于视觉-语言任务。
        .modelName("qwen-vl-max")
        .baseUrl("https://dashscope.aliyuncs.com/compatible-mode/v1")
        .build();
    }



    /**
     * @Description: 测试通义万象来实现图片生成，
     * 知识出处，https://help.aliyun.com/zh/model-studio/text-to-image
     * @Auther: zzyybs@126.com
     */
    @Bean
    public WanxImageModel wanxImageModel()
    {
        return WanxImageModel.builder()
        .apiKey("sk-cf4de7068032411b9de71adde80e3b12")
        .modelName("wanx2.1-t2i-turbo") //图片生成 https://help.aliyun.com/zh/model-studio/text-to-image
        .build();
    }
}

controller

方法createImageContent2和方法createImageContent3使用不同的方式访问大模型，其中createImageContent2为常规的使用langchain4j的api对大模型进行配置和提问，而createImageContent3使用dashscope原生的sdk对大模型进行配置和提问。区别在于抽象层次和使用的开发库不同。

特性createImageContent2 (使用 LangChain4j)createImageContent3 (使用原生 DashScope SDK)
核心库dev.langchain4j.community.model.dashscope.WanxImageModelcom.alibaba.dashscope.aigc.imagesynthesis.ImageSynthesis
抽象级别高层抽象 (High-Level Abstraction)底层原生 (Low-Level Native)
编程范式集成框架式调用。LangChain4j 作为一个AI应用开发框架，将底层API调用封装成了标准化的模型接口。开发者面向的是LangChain4j的WanxImageModel，而不是直接和阿里云的SDK打交道。原生SDK直接调用。直接使用阿里云官方提供的dashscope-sdk-java，对API的请求和响应进行精细化控制。
代码简洁度非常简洁。核心代码只有一行 wanxImageModel.generate("美女")。相对繁琐。需要手动构建一个详细的参数对象ImageSynthesisParam，包括模型名称、提示词、风格、数量、尺寸等。
配置方式外部化配置。API Key、模型名称、超时时间等参数通常通过Spring Boot的配置文件（如 application.yml）进行统一管理，然后通过@Autowired注入一个已经配置好的WanxImageModel实例。这符合企业级开发的实践。硬编码或方法内配置。在此示例中，API Key (sk-cf4de...)、模型 (WANX_V1)、风格 () 等参数直接在方法内部硬编码。这使得方法不灵活且不安全。
灵活性与控制力灵活性较低。LangChain4j提供的是通用参数，如果要使用某个模型非常独特且未被框架支持的参数，可能会比较困难。灵活性和控制力极高。可以直接设置原生SDK支持的所有参数，对API的控制力最强。
返回结果返回的是LangChain4j定义的标准Response对象。这样做的优点是，如果未来想从阿里云切换到其他厂商（如OpenAI DALL-E），代码改动会非常小，因为它们在LangChain4j中可能都返回同样类型的对象。返回的是DashScope SDK定义的ImageSynthesisResult对象。这个结果是JSON格式，包含了任务ID、状态、每个图片的URL等原生信息。
适用场景适用于快速构建和集成，或者在需要轻松切换不同模型提供商的复杂应用中。代码更优雅、可维护性更高。适用于需要对API进行深度定制和精细化控制的场景，或者在不使用大型框架的简单项目中直接调用。
import com.alibaba.dashscope.aigc.imagesynthesis.ImageSynthesis;
import com.alibaba.dashscope.aigc.imagesynthesis.ImageSynthesisParam;
import com.alibaba.dashscope.aigc.imagesynthesis.ImageSynthesisResult;
import com.alibaba.dashscope.exception.ApiException;
import com.alibaba.dashscope.exception.NoApiKeyException;
import com.alibaba.dashscope.utils.JsonUtils;
import dev.langchain4j.community.model.dashscope.WanxImageModel;
import dev.langchain4j.data.image.Image;
import dev.langchain4j.model.output.Response;
import lombok.extern.slf4j.Slf4j;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RestController;

import java.io.IOException;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-30 11:57
 * @Description: TODO
 */
@RestController
@Slf4j
public class WanxImageModelController
{
    @Autowired
    private WanxImageModel wanxImageModel;

    // http://localhost:9006/image/create2
    @GetMapping(value = "/image/create2")
    public String createImageContent2() throws IOException
    {
        System.out.println(wanxImageModel);
        Response<Image> imageResponse = wanxImageModel.generate("美女");

        System.out.println(imageResponse.content().url());

        return imageResponse.content().url().toString();

    }


    @GetMapping(value = "/image/create3")
    public String createImageContent3() throws IOException
    {

        String prompt = "近景镜头，18岁的中国女孩，古代服饰，圆脸，正面看着镜头，" +
                "民族优雅的服装，商业摄影，室外，电影级光照，半身特写，精致的淡妆，锐利的边缘。";
        ImageSynthesisParam param =
                ImageSynthesisParam.builder()
                            .apiKey("sk-cf4de7068032411b9de71adde80e3b12")
                            .model(ImageSynthesis.Models.WANX_V1)
                            .prompt(prompt)
                            .style("<watercolor>")
                            .n(1)
                            .size("1024*1024")
                        .build();

        ImageSynthesis imageSynthesis = new ImageSynthesis();
        ImageSynthesisResult result = null;

        try {
            System.out.println("---sync call, please wait a moment----");
            result = imageSynthesis.call(param);
        } catch (ApiException | NoApiKeyException e){
            throw new RuntimeException(e.getMessage());
        }


        System.out.println(JsonUtils.toJson(result));

        return JsonUtils.toJson(result);
    }
}

5.2.3. 测试

访问http://localhost:9006/image/create2测试createImageContent2

image

image

访问http://localhost:9006/image/create3测试测试createImageContent3

image

image

6. 流式输出

6.1.1. 说明

是一种逐步返回大模型生成结果的技术，生成一点返回一点，允许服务器将响应内容分批次实时传输给客户端，而不是等待全部内容生成完毕后再一次性返回。

流式输出的实现有两种方式

  1. low-level：docs.langchain4j.dev/tutorials/r…

image

image

  1. high-level 

image

6.1.2. 准备

pom

使用high-level需要新增依赖

<dependency>
    <groupId>dev.langchain4j</groupId>
    <artifactId>langchain4j-reactor</artifactId>
</dependency>

配置文件

设置响应的字符编码，避免流式返回输出乱码

server.servlet.encoding.charset=utf-8
server.servlet.encoding.enabled=true
server.servlet.encoding.force=true

6.1.3. 代码

servcie

high-level

import reactor.core.publisher.Flux;

public interface ChatAssistant
{
    //普通回答
    String chat(String prompt);
    //流式回答
    Flux<String> chatFlux(String prompt);
}

config

import com.atguigu.study.service.ChatAssistant;
import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.chat.StreamingChatModel;
import dev.langchain4j.model.openai.OpenAiChatModel;
import dev.langchain4j.model.openai.OpenAiStreamingChatModel;
import dev.langchain4j.service.AiServices;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;

import java.time.Duration;
import java.util.List;

@Configuration
public class LLMConfig
{

    @Bean(name = "qwen")
    public ChatModel chatModelQwen()
    {
        return OpenAiChatModel.builder()
                .apiKey(System.getenv("aliQwen-api"))
                .modelName("qwen-plus")
                .baseUrl("https://dashscope.aliyuncs.com/compatible-mode/v1")
                .build();
    }

    @Bean
    public StreamingChatModel streamingChatModel(){
        return OpenAiStreamingChatModel.builder()
                    .apiKey(System.getenv("aliQwen-api"))
                    .modelName("qwen-plus")
                    .baseUrl("https://dashscope.aliyuncs.com/compatible-mode/v1")
                .build();
    }

    @Bean
    public ChatAssistant chatAssistant(StreamingChatModel streamingChatModel){
        return AiServices.create(ChatAssistant.class, streamingChatModel);
    }
}

controller

import com.atguigu.study.service.ChatAssistant;
import dev.langchain4j.model.chat.StreamingChatModel;
import dev.langchain4j.model.chat.response.ChatResponse;
import dev.langchain4j.model.chat.response.StreamingChatResponseHandler;
import jakarta.annotation.Resource;
import jakarta.servlet.http.HttpServletResponse;
import lombok.extern.slf4j.Slf4j;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RequestParam;
import org.springframework.web.bind.annotation.RestController;
import reactor.core.publisher.Flux;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-30 16:25
 * @Description: 知识出处，https://docs.langchain4j.dev/tutorials/response-streaming
 */
@RestController
@Slf4j
public class StreamingChatModelController
{
    @Resource //直接使用 low-level LLM API
    private StreamingChatModel streamingChatLanguageModel;
    @Resource //自己封装接口使用 high-level LLM API
    private ChatAssistant chatAssistant;


    // http://localhost:9007/chatstream/chat?prompt=天津有什么好吃的
    @GetMapping(value = "/chatstream/chat")
    public Flux<String> chat(@RequestParam("prompt") String prompt)
    {
        System.out.println("---come in chat");

        return Flux.create(emitter -> {
            streamingChatLanguageModel.chat(prompt, new StreamingChatResponseHandler()
            {
                @Override
                public void onPartialResponse(String partialResponse)
                {
                    emitter.next(partialResponse);
                }

                @Override
                public void onCompleteResponse(ChatResponse completeResponse)
                {
                    emitter.complete();
                }

                @Override
                public void onError(Throwable throwable)
                {
                    emitter.error(throwable);
                }
            });
        });
    }

    @GetMapping(value = "/chatstream/chat3")
    public Flux<String> chat3(@RequestParam(value = "prompt", defaultValue = "南京有什么好吃") String prompt)
    {
        System.out.println("---come in chat3");

        return chatAssistant.chatFlux(prompt);
    }
}

6.1.4. 测试

访问http://localhost:9007/chatstream/chat?prompt=天津有什么好吃的测试low-level

image

image

访问http://localhost:9007/chatstream/chat3 测试high-level

image

image

7. 记忆缓存

上下文记忆是用户在和模型进行多轮对话时，模型有记忆上下文对话内容的功能。从而提供连贯和个性化的回复。

image

7.1. 说明

image

image

注意，使用TokenWindowChatMemory选项是需要结合Tokenizer计算token数量，这个计算过程会耗费一定的时间，尤其是在历史记录很长的情况下。

由于我们要测验记忆功能，考虑到问答上下文很长，我们换一个能支持更大上下文长度的模型qwen-long

image.png

7.2. 准备

MessageWindowChatMemory基于消息数量的简单实现。它采用滑动窗口的方式，保留最近的n条数据并清楚旧消息。

7.3. 代码（基于MessageWindowChatMemory）

config

这里我们使用high-level的方式，并且规定消息的最大长度是100，当消息数量超过100时，会触发清除旧消息的机制。

import com.atguigu.study.service.ChatAssistant;
import com.atguigu.study.service.ChatMemoryAssistant;
import dev.langchain4j.memory.chat.MessageWindowChatMemory;
import dev.langchain4j.memory.chat.TokenWindowChatMemory;
import dev.langchain4j.model.TokenCountEstimator;
import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.openai.OpenAiChatModel;
import dev.langchain4j.model.openai.OpenAiTokenCountEstimator;
import dev.langchain4j.service.AiServices;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;

@Configuration
public class LLMConfig
{
    @Bean
    public ChatModel chatModel()
    {
        return OpenAiChatModel.builder()
                    .apiKey("sk-cf4de7068032411b9de71adde80e3b12")
                    .modelName("qwen-long")
                    .baseUrl("https://dashscope.aliyuncs.com/compatible-mode/v1")
                .build();
    }

    @Bean(name = "chat")
    public ChatAssistant chatAssistant(ChatModel chatModel)
    {
        return AiServices.create(ChatAssistant.class, chatModel);
    }

    @Bean(name = "chatMessageWindowChatMemory")
    public ChatMemoryAssistant chatMessageWindowChatMemory(ChatModel chatModel)
    {

        return AiServices.builder(ChatMemoryAssistant.class)
                .chatModel(chatModel)
                //按照memoryId对应创建了一个chatMemory
                .chatMemoryProvider(memoryId -> MessageWindowChatMemory.withMaxMessages(100))
                .build();
    }

}

controller

方法 chat()为没有记忆缓存功能，方法chatMessageWindowChatMemory()是有记忆缓存功能

import cn.hutool.core.date.DateUtil;
import com.atguigu.study.service.ChatAssistant;
import com.atguigu.study.service.ChatMemoryAssistant;
import dev.langchain4j.data.message.AiMessage;
import dev.langchain4j.data.message.UserMessage;
import dev.langchain4j.model.chat.response.ChatResponse;
import jakarta.annotation.Resource;
import lombok.extern.slf4j.Slf4j;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RestController;

import java.util.Arrays;

@RestController
@Slf4j
public class ChatMemoryController
{
    @Resource(name = "chat")
    private ChatAssistant chatAssistant;

    @Resource(name = "chatMessageWindowChatMemory")
    private ChatMemoryAssistant chatMessageWindowChatMemory;

    @Resource(name = "chatTokenWindowChatMemory")
    private ChatMemoryAssistant chatTokenWindowChatMemory;

    @GetMapping(value = "/chatmemory/test1")
    public String chat()
    {

        String answer01 = chatAssistant.chat("你好，我的名字叫张三");
        System.out.println("answer01返回结果："+answer01);

        String answer02 = chatAssistant.chat("我的名字是什么");
        System.out.println("answer02返回结果："+answer02);

        return "success : "+ DateUtil.now()+"<br> \n\n answer01: "+answer01+"<br> \n\n answer02: "+answer02;
    }

    @GetMapping(value = "/chatmemory/test2")
    public String chatMessageWindowChatMemory()
    {
        chatMessageWindowChatMemory.chatWithChatMemory(1L, "你好！我的名字是Java.");
        String answer01 = chatMessageWindowChatMemory.chatWithChatMemory(1L, "我的名字是什么");
        System.out.println("answer01返回结果："+answer01);

        chatMessageWindowChatMemory.chatWithChatMemory(3L, "你好！我的名字是C++");
        String answer02 = chatMessageWindowChatMemory.chatWithChatMemory(3L, "我的名字是什么");
        System.out.println("answer02返回结果："+answer02);

        return "chatMessageWindowChatMemory success : "
                + DateUtil.now()+"<br> \n\n answer01: "+answer01+"<br> \n\n answer02: "+answer02;

    }
}

service

普通没有记忆功能的service。

import dev.langchain4j.service.MemoryId;
import dev.langchain4j.service.UserMessage;

public interface ChatAssistant
{
    /**
    * @Description: 普通聊天对话，不带记忆缓存功能
    */
    String chat(String prompt);
}

有记忆功能的service。

import dev.langchain4j.service.MemoryId;
import dev.langchain4j.service.UserMessage;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-30 19:22
 * @Description: TODO
 */
public interface ChatMemoryAssistant
{

    /**
     * 聊天带记忆缓存功能
     *
     * @param userId  用户 ID
     * @param prompt 消息
     * @return {@link String }
     */
    String chatWithChatMemory(@MemoryId Long userId, @UserMessage String prompt);
}

7.4. 测试

访问http://localhost:9008/chatmemory/test1 测试无记忆功能的接口

image.png

访问http://localhost:9008/chatmemory/test2 测试基于MessageWindowChatMemory有记忆功能的接口

image.png

7.5. 代码（基于TokenWindowChatMemory）

config

这里我们使用high-level的方式，规定消息最多1000个token。

import com.atguigu.study.service.ChatAssistant;
import com.atguigu.study.service.ChatMemoryAssistant;
import dev.langchain4j.memory.chat.MessageWindowChatMemory;
import dev.langchain4j.memory.chat.TokenWindowChatMemory;
import dev.langchain4j.model.TokenCountEstimator;
import dev.langchain4j.model.chat.ChatModel;
import dev.langchain4j.model.openai.OpenAiChatModel;
import dev.langchain4j.model.openai.OpenAiTokenCountEstimator;
import dev.langchain4j.service.AiServices;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;

/**
 * @auther zzyybs@126.com
 * @Date 2025-05-30 18:58
 * @Description: 知识出处，https://docs.langchain4j.dev/tutorials/chat-memory/#eviction-policy
 */
@Configuration
public class LLMConfig
{
    @Bean(name = "chatTokenWindowChatMemory")
    public ChatMemoryAssistant chatTokenWindowChatMemory(ChatModel chatModel)
    {
        //1 TokenCountEstimator默认的token分词器，需要结合Tokenizer计算ChatMessage的token数量
        TokenCountEstimator openAiTokenCountEstimator = new OpenAiTokenCountEstimator("gpt-4");

        return AiServices.builder(ChatMemoryAssistant.class)
                .chatModel(chatModel)
                .chatMemoryProvider(memoryId -> TokenWindowChatMemory.withMaxTokens(1000,openAiTokenCountEstimator))
                .build();
    }
}

controller

import cn.hutool.core.date.DateUtil;
import com.atguigu.study.service.ChatAssistant;
import com.atguigu.study.service.ChatMemoryAssistant;
import dev.langchain4j.data.message.AiMessage;
import dev.langchain4j.data.message.UserMessage;
import dev.langchain4j.model.chat.response.ChatResponse;
import jakarta.annotation.Resource;
import lombok.extern.slf4j.Slf4j;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RestController;

import java.util.Arrays;

@RestController
@Slf4j
public class ChatMemoryController
{
    @Resource(name = "chat")
    private ChatAssistant chatAssistant;

    @Resource(name = "chatMessageWindowChatMemory")
    private ChatMemoryAssistant chatMessageWindowChatMemory;

    @Resource(name = "chatTokenWindowChatMemory")
    private ChatMemoryAssistant chatTokenWindowChatMemory;

    @GetMapping(value = "/chatmemory/test3")
    public String chatTokenWindowChatMemory()
    {
        chatTokenWindowChatMemory.chatWithChatMemory(1L, "你好！我的名字是mysql");
        String answer01 = chatTokenWindowChatMemory.chatWithChatMemory(1L, "我的名字是什么");
        System.out.println("answer01返回结果："+answer01);

        chatTokenWindowChatMemory.chatWithChatMemory(3L, "你好！我的名字是oracle");
        String answer02 = chatTokenWindowChatMemory.chatWithChatMemory(3L, "我的名字是什么");
        System.out.println("answer02返回结果："+answer02);

        return "chatTokenWindowChatMemory success : "
                + DateUtil.now()+"<br> \n\n answer01: "+answer01+"<br> \n\n answer02: "+answer02;
    }
}

service同上

7.5. 测试

访问http://localhost:9008/chatmemory/test3测试

image.png

avatar
文章
阅读
粉丝