数据仓库与分析SpringAI调用OpenAI Demo

诗林发表于 2024-6-20 22:43:46

SpringAI调用OpenAI Demo

Spring AI

https://img-blog.csdnimg.cn/direct/87cc497bd26d42c68c8d5d61a5ad57bd.png

在maven的setting.xml
<mirror>
<id>spring-milestones</id>
<name>Spring Milestones</name>
<mirrorOf>spring-milestones</mirrorOf>
<url>https://repo.spring.io/milestone</url>
</mirror>
这里以调用GPT4o为例
https://img-blog.csdnimg.cn/direct/e0345034ba2e46509b469d60cfc88fc6.png
https://img-blog.csdnimg.cn/direct/fe3d282f0fa9468e9adbac562ef05267.png

厥后为了测试JDK8是否可用将版本调整成了2.7.2 结果不能使用
https://img-blog.csdnimg.cn/direct/b9270341c1a84e01b9ddd4941dcb52ea.png

因国内无法直接访问按了个nginx代理
https://img-blog.csdnimg.cn/direct/f0c2db285f9945b5a8b3e85055c2feed.png

server {
#HTTPS的默认访问端口443。
#如果未在此处配置HTTPS的默认访问端口，可能会造成Nginx无法启动。
listen 443 ssl;

#填写证书绑定的域名
server_name xxxx xxxxxx;

#填写证书文件绝对路径
ssl_certificate /etc/letsencrypt/live/xxx.com/fullchain.pem;
#填写证书私钥文件绝对路径
ssl_certificate_key /etc/letsencrypt/live/xxxx.com/privkey.pem;

ssl_session_cache shared:SSL:1m;
ssl_session_timeout 5m;

#自定义设置使用的TLS协议的类型以及加密套件（以下为配置示例，请您自行评估是否需要配置）
#TLS协议版本越高，HTTPS通信的安全性越高，但是相较于低版本TLS协议，高版本TLS协议对浏览器的兼容性较差。
ssl_ciphers ECDHE-RSA-AES128-GCM-SHA256:ECDHE:ECDH:AES:HIGH:!NULL:!aNULL:!MD5:!ADH:!RC4;
#ssl_protocols TLSv1.1 TLSv1.2 TLSv1.3;

#表示优先使用服务端加密套件。默认开启
ssl_prefer_server_ciphers on;
location /v1/{
         chunked_transfer_encoding off;
         proxy_cache off;
         proxy_buffering off;
         proxy_redirect off;
         proxy_ssl_protocols TLSv1 TLSv1.1 TLSv1.2;
         proxy_ssl_server_name on;
         proxy_http_version 1.1;
         proxy_set_header Host api.openai.com;
proxy_set_header X-Real-IP $server_addr;
proxy_set_header X-Forwarded-For $server_addr;
proxy_set_header X-Real-Port $server_port;
         proxy_set_header Connection '';
proxy_pass https://api.openai.com/;

   }

https://img-blog.csdnimg.cn/direct/c60dc75c403f48f1a8d306a22dc796b7.png

配置ChatClient另外种方式
https://img-blog.csdnimg.cn/direct/b4397a5886f84bf6a25bdaaa3ce98b4b.png
https://img-blog.csdnimg.cn/direct/ea1e7f6591744482884d054223bbcda8.png

特定的对话风格或角色，通常发起详细定义你盼望模子如何回应，然后在你的应用中相应地构建提示其实就是对话之前
https://img-blog.csdnimg.cn/direct/fb062825532844c7b1288f03d831dba1.png
https://img-blog.csdnimg.cn/direct/b0d27981b6b34de7bb6b9216dfda56ea.png

https://img-blog.csdnimg.cn/direct/3ae4c32f93e84bd9b514576d1fd3cf49.png
https://img-blog.csdnimg.cn/direct/0753018019954493af9ad97f79d8abc6.png

你也可以为每一个接口设置单独的预定义角色比方
https://img-blog.csdnimg.cn/direct/226261f7c9a9461f9317feda62fc680f.png

以流的方式返回
https://img-blog.csdnimg.cn/direct/190efff58f624b1aaaa3fe09e24c5525.png

这个在postMan中不好提现

可以直接在浏览器

可以看到它是以流的方式返回的,但是乱码
https://img-blog.csdnimg.cn/direct/b72bde27db51405b8fd098428a486bc1.png
https://img-blog.csdnimg.cn/direct/b274020e9c1e4cee905809c754c69dd5.png

https://img-blog.csdnimg.cn/direct/e914ff76245b459483526bcdcf4e5ee2.png

除了使用nginx转发还可以用本地代理只要在应用启动前配置好就行
https://img-blog.csdnimg.cn/direct/c59ce48b7a064fb1997aa5dbcb4d2555.png
https://img-blog.csdnimg.cn/direct/05e54326eeb04d9593d251811cb6a459.png

关于ChatClient和ChatModel
ChatClient:较为通用
ChatModel:设置模子独有功能
https://img-blog.csdnimg.cn/direct/5f265c3a98fc43b28f9ee4c35b0c714b.png
https://img-blog.csdnimg.cn/direct/20991fb428794ec88010a8f79ba12121.png
模子选择
https://img-blog.csdnimg.cn/direct/11560fba970c4138b81c05b03ab4e32e.png
https://img-blog.csdnimg.cn/direct/a4b1bf62c9f24750a8af3c1c7443fed8.png
https://img-blog.csdnimg.cn/direct/ed581ba910fe44a2848e4a848b0af138.png

https://img-blog.csdnimg.cn/direct/94fe8d9344d848cbb883b32c25da2b83.png
下面使用ChatModel演示调用
可以参数中指定, 也可以application.properties中指定
https://img-blog.csdnimg.cn/direct/45e302068f2946b1b7b70a984273d4bd.png

流式
https://img-blog.csdnimg.cn/direct/2dcb2aa5b5e248549982e630811a126c.png

演示文生图功能
https://img-blog.csdnimg.cn/direct/10b3e1428cfb47e3b91479290c0e11ba.png

https://img-blog.csdnimg.cn/direct/5de2628076184aadac2379f6a0f6afc9.png
https://img-blog.csdnimg.cn/direct/8a515cac634a46c285f630858aa07070.png
https://img-blog.csdnimg.cn/direct/be94f2aefa274734975d917562e8abda.png

文生语音
https://img-blog.csdnimg.cn/direct/45718f199ad047c9b1cd9b84c59a77da.png
下面做法是有题目的,由于你保存到resources目次下的话项目是打包之后运行的因此你第一次运行保存之后是读不到的要读只能重新启动,这里只是演示就先如许了
https://img-blog.csdnimg.cn/direct/dfb2402a684046c9af0f8ce56bc1d06f.png
https://img-blog.csdnimg.cn/direct/0a5bd2d9d6e64185831ff70cb25e6d22.png

重启应用
https://img-blog.csdnimg.cn/direct/e11e5717f3fc4fe9a91c3fd7b3038e61.png

关于语音转文本
https://img-blog.csdnimg.cn/direct/5d26b7bab03143f5834628b151dbdc78.png

https://img-blog.csdnimg.cn/direct/22b5011a58544be6811f00e224a5cfce.png
https://img-blog.csdnimg.cn/direct/3ab62967789f4041894c9363ef9cd818.png

关于多模态(意思就是你可以要发文本,要发图片,要发语音)
https://img-blog.csdnimg.cn/direct/665eba01018b449ba05c53f902c9b5cc.png

意思只能用GPT4或4o模子才气用多模态
https://img-blog.csdnimg.cn/direct/afe88831ebfc4fc7bd5b28c4dc919144.png
https://img-blog.csdnimg.cn/direct/24518d850e5042adb535fb5a578fe5c8.png

https://img-blog.csdnimg.cn/direct/0ed58fda2e124bd592d9f9e5f26a74e0.png

https://img-blog.csdnimg.cn/direct/8a35822edea04c34881127805c8288a4.png

以上的代码
package com.example.springai.controller;

import org.springframework.ai.chat.client.ChatClient;
import org.springframework.ai.chat.messages.Media;
import org.springframework.ai.chat.messages.UserMessage;
import org.springframework.ai.chat.model.ChatResponse;
import org.springframework.ai.chat.prompt.Prompt;
import org.springframework.ai.image.ImagePrompt;
import org.springframework.ai.image.ImageResponse;
import org.springframework.ai.openai.*;
import org.springframework.ai.openai.api.OpenAiApi;
import org.springframework.ai.openai.api.OpenAiAudioApi;
import org.springframework.ai.openai.audio.speech.SpeechPrompt;
import org.springframework.ai.openai.audio.speech.SpeechResponse;
import org.springframework.ai.openai.audio.transcription.AudioTranscriptionPrompt;
import org.springframework.ai.openai.audio.transcription.AudioTranscriptionResponse;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.core.io.ClassPathResource;
import org.springframework.http.HttpStatus;
import org.springframework.http.ResponseEntity;
import org.springframework.util.MimeTypeUtils;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RestController;
import reactor.core.publisher.Flux;

import java.io.IOException;
import java.nio.file.Files;
import java.nio.file.Path;
import java.nio.file.Paths;
import java.util.List;

/**
* @author hrui
* @date 2024/6/8 2:19
*/
@RestController
public class HelloGPT {
@Autowired
private ChatClient chatClient;

// public HelloGPT(ChatClient.Builder chatClientBuilder) {
//    this.chatClient=chatClientBuilder.build();
// }

@GetMapping("/helloai")
public Object generate(String userInput) {
   System.out.println("userInput:"+userInput);
   return chatClient.prompt()//提示词
            .user(userInput)//用户输入
            //.system("You are a helpful assistant.")
            .call()//调用
            .content();//返回文本
}

@GetMapping(value = "/helloai2",produces = "text/html;charset=UTF-8")
public Flux<String> generate2(String userInput) {
   Flux<String> output = chatClient.prompt()
            .user(userInput)
            .stream()
            .content();
   return output;
}

@Autowired
private OpenAiChatModel chatModel;//ChatModel可以自动装配不需要@Bean

@GetMapping("/helloai3")
public Object generate3(String userInput) {
//    ChatResponse response = chatModel.call(
//             new Prompt(
//                      "Generate the names of 5 famous pirates.",//这个其实好比用户消息
//                      OpenAiChatOptions.builder()
//                            .withModel("gpt-4-32k")
//                            .withTemperature(0.8F)
//                            .build()
//             ));
   ChatResponse response = chatModel.call(
            new Prompt(
                     userInput,//底层封装成new UserMessage(userInput)
                     OpenAiChatOptions.builder()
                           .withModel("gpt-4-turbo")
                           .withTemperature(0.8F)
                           .build()
            ));
   return response.getResult().getOutput().getContent();
}

@GetMapping("/helloai4")
public Flux<ChatResponse> generate4(String userInput) {
   System.out.println("userInput:"+userInput);
   Flux<ChatResponse> stream = chatModel.stream(
            new Prompt(
                     userInput//底层封装成new UserMessage(userInput)

            ));

   return stream;
}

@Autowired
private OpenAiImageModel openAiImageModel;

@GetMapping("/helloai6")
public Object generate6(String userInput) {
   ImageResponse response = openAiImageModel.call(
            new ImagePrompt(userInput,
                     OpenAiImageOptions.builder()
                           //设置图片清晰度
                           .withQuality("hd")
                           .withModel("dall-e-3")//默认就是这个
                           .withN(1)//生成几张图片
                           //默认高度和宽度
                           .withHeight(1024)
                           .withWidth(1024).build())

   );

   return response.getResult().getOutput().getUrl();
}

// @Autowired
// private OpenAiAudioTranscriptionModel openAiAudioTranscriptionModel;
@Autowired
private OpenAiAudioSpeechModel openAiAudioSpeechModel;

@GetMapping("/helloai7")
public Object generate7(String userInput) {
   OpenAiAudioSpeechOptions speechOptions = OpenAiAudioSpeechOptions.builder()
            //用的模型
            .withModel(OpenAiAudioApi.TtsModel.TTS_1.value)
            //设置人声
            .withVoice(OpenAiAudioApi.SpeechRequest.Voice.ALLOY)
            .withResponseFormat(OpenAiAudioApi.SpeechRequest.AudioResponseFormat.MP3)
            .withSpeed(1.0f)//合成语音的速度 0.0最慢1.0最快
            .build();

   SpeechPrompt speechPrompt = new SpeechPrompt(userInput, speechOptions);
   SpeechResponse response = openAiAudioSpeechModel.call(speechPrompt);

   byte[] output = response.getResult().getOutput();

   try {
         // 指定文件名，这里以当前时间戳命名以避免重名
         String filename = "audio_" + System.currentTimeMillis() + ".mp3";
         // 指定保存路径
         Path path = Paths.get("src/main/resources/static/" + filename);
         // 写入文件
         Files.write(path, output);

         // 获取可访问的URL，假设你的服务运行在 localhost:8080
         String fileUrl = "http://localhost:8082/" + filename;
         return ResponseEntity.ok(fileUrl);
   } catch (Exception e) {
         e.printStackTrace();
         return ResponseEntity.status(HttpStatus.INTERNAL_SERVER_ERROR).body("Error saving file");
   }

}

@Autowired
private OpenAiAudioTranscriptionModel openAiTranscriptionModel;

@GetMapping("/helloai8")
public Object generate8() {
   //语音翻译的可选配置
   var transcriptionOptions = OpenAiAudioTranscriptionOptions.builder()
            .withResponseFormat(OpenAiAudioApi.TranscriptResponseFormat.TEXT)
            //温度0f不需要创造力,语音是什么,翻译什么
            .withTemperature(0f)
            .build();
   var audioFile=new ClassPathResource("hello.m4a");
   //var audioFile = new FileSystemResource("/path/to/your/resource/speech/jfk.flac");

   AudioTranscriptionPrompt transcriptionRequest = new AudioTranscriptionPrompt(audioFile, transcriptionOptions);
   AudioTranscriptionResponse response = openAiTranscriptionModel.call(transcriptionRequest);

   return response.getResult().getOutput();
}

@GetMapping("/helloai9")
public Object generate9() throws IOException {
   //图片二进制流
   byte[] imageData=new ClassPathResource("1717865484569.png").getContentAsByteArray();

   //用户信息
   var userMessage=new UserMessage("这是什么", List.of(new Media(MimeTypeUtils.IMAGE_PNG,imageData)));

   OpenAiChatOptions build = OpenAiChatOptions.builder()
            .withModel(OpenAiApi.ChatModel.GPT_4_O.getValue()).build();

   ChatResponse response=chatModel.call(new Prompt(userMessage,build));

   return response.getResult().getOutput().getContent();
}
}

关于Function call 应对大模子无法获取实时信息的毛病
比如说,我现在问本日杭州火车东站的客流量是多少,如许GPT肯定无法回答
那么需要怎么办呢我们可以调用第三方接口得到信息再告知GPT 然后GPT回答题目
https://img-blog.csdnimg.cn/direct/6bc5ea881b8c480187c90d48d0f26365.png
https://img-blog.csdnimg.cn/direct/3a616dfed9be432aa95544736fe78bd2.png

大概表明
比方我问杭州有多少生齿
这类题目,GPT是无法回答的,当然现在GPT会查阅相关资料回答,假设
这句话里有 location和count两个关键
Function Call的作用是当问GPT一个类似题目之后,GPT用Function Call往返调我们的应用并携带关键信息 location和count信息,我们的应用去查数据库也好,去调用第三方接口也好,再告诉GPT 那么GPT就可以回答这个题目了

当GPT携带参数过来的时间会调用Function.apply(){}这个方法,那么我们在这个方法里写我们自己的逻辑可以查数据库,可以调用第三方接口

创建一个实现Function接口的类
https://img-blog.csdnimg.cn/direct/a845e1b6dee04f0582f5d222967654ea.png

如果报错
https://img-blog.csdnimg.cn/direct/30ba686109fd4fceb56b4974482f362d.png

https://img-blog.csdnimg.cn/direct/5f607802473643e990892fecc7141dd7.png
https://img-blog.csdnimg.cn/direct/405f926838a04facb9ecdc625a610b26.png
https://img-blog.csdnimg.cn/direct/963dbf166a8e418fab3f8ec9b18124ba.png

全部代码
package com.example.springai.controller;

import org.springframework.ai.chat.model.ChatResponse;
import org.springframework.ai.chat.prompt.Prompt;
import org.springframework.ai.openai.OpenAiChatModel;
import org.springframework.ai.openai.OpenAiChatOptions;
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;

/**
* @author hrui
* @date 2024/6/9 1:16
*/
@RestController
public class FunctionCallController {

@Autowired
private OpenAiChatModel chatModel;

@GetMapping("/functionCall")
public Object functionCall(@RequestParam(value = "message",defaultValue = "杭州有多少个名字叫韩妹妹的人") String message) {

   OpenAiChatOptions aiChatOptions=OpenAiChatOptions.builder()
            //设置实现了Function接口的beanName
   .withFunction("locationCount")
   .withModel("gpt-3.5-turbo")
   .build();

   ChatResponse response=chatModel.call(new Prompt(message,aiChatOptions));
   //Flux<ChatResponse> stream = chatModel.stream(new Prompt(message, aiChatOptions));
   return response.getResult().getOutput().getContent();
}
}

package com.example.springai.controller;

import org.springframework.ai.chat.client.ChatClient;
import org.springframework.ai.chat.messages.Media;
import org.springframework.ai.chat.messages.UserMessage;
import org.springframework.ai.chat.model.ChatResponse;
import org.springframework.ai.chat.prompt.Prompt;
import org.springframework.ai.image.ImagePrompt;
import org.springframework.ai.image.ImageResponse;
import org.springframework.ai.openai.*;
import org.springframework.ai.openai.api.OpenAiApi;
import org.springframework.ai.openai.api.OpenAiAudioApi;
import org.springframework.ai.openai.audio.speech.SpeechPrompt;
import org.springframework.ai.openai.audio.speech.SpeechResponse;
import org.springframework.ai.openai.audio.transcription.AudioTranscriptionPrompt;
import org.springframework.ai.openai.audio.transcription.AudioTranscriptionResponse;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.core.io.ClassPathResource;
import org.springframework.http.HttpStatus;
import org.springframework.http.ResponseEntity;
import org.springframework.util.MimeTypeUtils;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RestController;
import reactor.core.publisher.Flux;

import java.io.IOException;
import java.nio.file.Files;
import java.nio.file.Path;
import java.nio.file.Paths;
import java.util.List;

/**
* @author hrui
* @date 2024/6/8 2:19
*/
@RestController
public class HelloGPT {
@Autowired
private ChatClient chatClient;

// public HelloGPT(ChatClient.Builder chatClientBuilder) {
//    this.chatClient=chatClientBuilder.build();
// }

@GetMapping("/helloai")
public Object generate(String userInput) {
   System.out.println("userInput:"+userInput);
   return chatClient.prompt()//提示词
            .user(userInput)//用户输入
            //.system("You are a helpful assistant.")
            .call()//调用
            .content();//返回文本
}

@GetMapping(value = "/helloai2",produces = "text/html;charset=UTF-8")
public Flux<String> generate2(String userInput) {
   Flux<String> output = chatClient.prompt()
            .user(userInput)
            .stream()
            .content();
   return output;
}

@Autowired
private OpenAiChatModel chatModel;//ChatModel可以自动装配不需要@Bean

@GetMapping("/helloai3")
public Object generate3(String userInput) {
//    ChatResponse response = chatModel.call(
//             new Prompt(
//                      "Generate the names of 5 famous pirates.",//这个其实好比用户消息
//                      OpenAiChatOptions.builder()
//                            .withModel("gpt-4-32k")
//                            .withTemperature(0.8F)
//                            .build()
//             ));
   ChatResponse response = chatModel.call(
            new Prompt(
                     userInput,//底层封装成new UserMessage(userInput)
                     OpenAiChatOptions.builder()
                           .withModel("gpt-4-turbo")
                           .withTemperature(0.8F)
                           .build()
            ));
   return response.getResult().getOutput().getContent();
}

@GetMapping("/helloai4")
public Flux<ChatResponse> generate4(String userInput) {
   System.out.println("userInput:"+userInput);
   Flux<ChatResponse> stream = chatModel.stream(
            new Prompt(
                     userInput//底层封装成new UserMessage(userInput)

            ));

   return stream;
}

@Autowired
private OpenAiImageModel openAiImageModel;

@GetMapping("/helloai6")
public Object generate6(String userInput) {
   ImageResponse response = openAiImageModel.call(
            new ImagePrompt(userInput,
                     OpenAiImageOptions.builder()
                           //设置图片清晰度
                           .withQuality("hd")
                           .withModel("dall-e-3")//默认就是这个
                           .withN(1)//生成几张图片
                           //默认高度和宽度
                           .withHeight(1024)
                           .withWidth(1024).build())

   );

   return response.getResult().getOutput().getUrl();
}

// @Autowired
// private OpenAiAudioTranscriptionModel openAiAudioTranscriptionModel;
@Autowired
private OpenAiAudioSpeechModel openAiAudioSpeechModel;

@GetMapping("/helloai7")
public Object generate7(String userInput) {
   OpenAiAudioSpeechOptions speechOptions = OpenAiAudioSpeechOptions.builder()
            //用的模型
            .withModel(OpenAiAudioApi.TtsModel.TTS_1.value)
            //设置人声
            .withVoice(OpenAiAudioApi.SpeechRequest.Voice.ALLOY)
            .withResponseFormat(OpenAiAudioApi.SpeechRequest.AudioResponseFormat.MP3)
            .withSpeed(1.0f)//合成语音的速度 0.0最慢1.0最快
            .build();

   SpeechPrompt speechPrompt = new SpeechPrompt(userInput, speechOptions);
   SpeechResponse response = openAiAudioSpeechModel.call(speechPrompt);

   byte[] output = response.getResult().getOutput();

   try {
         // 指定文件名，这里以当前时间戳命名以避免重名
         String filename = "audio_" + System.currentTimeMillis() + ".mp3";
         // 指定保存路径
         Path path = Paths.get("src/main/resources/static/" + filename);
         // 写入文件
         Files.write(path, output);

         // 获取可访问的URL，假设你的服务运行在 localhost:8080
         String fileUrl = "http://localhost:8082/" + filename;
         return ResponseEntity.ok(fileUrl);
   } catch (Exception e) {
         e.printStackTrace();
         return ResponseEntity.status(HttpStatus.INTERNAL_SERVER_ERROR).body("Error saving file");
   }

}

@Autowired
private OpenAiAudioTranscriptionModel openAiTranscriptionModel;

@GetMapping("/helloai8")
public Object generate8() {
   //语音翻译的可选配置
   var transcriptionOptions = OpenAiAudioTranscriptionOptions.builder()
            .withResponseFormat(OpenAiAudioApi.TranscriptResponseFormat.TEXT)
            //温度0f不需要创造力,语音是什么,翻译什么
            .withTemperature(0f)
            .build();
   var audioFile=new ClassPathResource("hello.m4a");
   //var audioFile = new FileSystemResource("/path/to/your/resource/speech/jfk.flac");

   AudioTranscriptionPrompt transcriptionRequest = new AudioTranscriptionPrompt(audioFile, transcriptionOptions);
   AudioTranscriptionResponse response = openAiTranscriptionModel.call(transcriptionRequest);

   return response.getResult().getOutput();
}

@GetMapping("/helloai9")
public Object generate9() throws IOException {
   //图片二进制流
   byte[] imageData=new ClassPathResource("1717865484569.png").getContentAsByteArray();

   //用户信息
   var userMessage=new UserMessage("这是什么", List.of(new Media(MimeTypeUtils.IMAGE_PNG,imageData)));

   OpenAiChatOptions build = OpenAiChatOptions.builder()
            .withModel(OpenAiApi.ChatModel.GPT_4_O.getValue()).build();

   ChatResponse response=chatModel.call(new Prompt(userMessage,build));

   return response.getResult().getOutput().getContent();
}
}

package com.example.springai.function;

import org.springframework.context.annotation.Description;
import org.springframework.stereotype.Component;
import org.springframework.util.StringUtils;

import java.util.function.Function;

/**
* @author hrui
* @date 2024/6/9 1:46
*/
@Component("locationCount")
@Description("某个地方有多少个名字某某的人")
public class LocationCount implements Function<LocationCount.Request, LocationCount.Response> {

@Override//接收GPT提取后的信息(自动调用该方法)
public Response apply(Request request) {
   //好的(聪明的模型)当location和name确实时候根本不会来调用该方法
   if(!StringUtils.hasLength(request.location)||!StringUtils.hasLength(request.name)){
         return new Response("参数缺失,无需Function-Call,正常响应就可以");
   }
   //如果有值调用数据库或者第三方接口

   return new Response("只有唯一强调唯一1个");
}

//密封类(好比有了@Data作用) 接收GPT信息
public record Request(String name, String location) {}
//密封类,最终响应GPT信息
public record Response(String message) {}

}

package com.example.springai;

import org.springframework.ai.chat.client.ChatClient;
import org.springframework.ai.chat.model.ChatModel;
import org.springframework.boot.SpringApplication;
import org.springframework.boot.autoconfigure.SpringBootApplication;
import org.springframework.context.annotation.Bean;

@SpringBootApplication
public class SpringaiApplication {

@Bean
public ChatClient chatClient(ChatClient.Builder chatClientBuilder){
   return chatClientBuilder.defaultSystem("你现在不是ChatGPT,也不是人工智能AI,现在开始你是一个正常人类,你的名字叫hrui")
            .build();
}
public static void main(String[] args) {
   //System.setProperty("proxyType","4");
   System.setProperty("proxyHost","127.0.0.1");
   System.setProperty("proxyPort","7897");
   //System.setProperty("proxySet","true");
   SpringApplication.run(SpringaiApplication.class, args);
}

}

免责声明：如果侵犯了您的权益，请联系站长，我们会及时删除侵权内容，谢谢合作！更多信息从访问主页：qidao123.com:ToB企服之家，中国第一个企服评测及商务社交产业平台。

页: [1]

ToB企服应用市场:ToB评测及商务社交产业平台's Archiver

SpringAI调用OpenAI Demo