java简单访问chatapt

java简单调用 gpt-3 api

引入依赖

1. 进入官方推荐的 java sdk
   • 文档介绍使用openai可以使用api、client和service三种包，service最简单，本次只使用service。
2. 找到依赖信息.
   implementation ‘com.theokanning.openai-gpt3-java::’
3. 在maven库中找到所需版本，在浏览器搜索
   mvn com.theokanning.openai-gpt3-java
4. 找到service包的所需配置代码
   • gradle

    implementation 'com.theokanning.openai-gpt3-java:service:0.10.0'
    
    //https://mvnrepository.com/artifact/com.fasterxml.jackson.core/jackson-core  
    implementation 'com.fasterxml.jackson.core:jackson-core:2.13.4'
   
    //https://mvnrepository.com/artifact/com.fasterxml.jackson.core/jackson-databind
    implementation 'com.fasterxml.jackson.core:jackson-databind:2.13.4.2'
   
   //https://mvnrepository.com/artifact/com.fasterxml.jackson.core/jackson-annotations
   implementation 'com.fasterxml.jackson.core:jackson-annotations:2.13.4'
   

   • maven

       <dependency>
          <groupId>com.theokanning.openai-gpt3-javagroupId>
          <artifactId>serviceartifactId>
          <version>0.10.0version>
       dependency> 
   
    
       <dependency>
          <groupId>com.fasterxml.jackson.coregroupId>
          <artifactId>jackson-coreartifactId>
          <version>2.13.4version>
       dependency>
    
       
       <dependency>
          <groupId>com.fasterxml.jackson.coregroupId>
          <artifactId>jackson-databindartifactId>
          <version>2.13.4.2version>
       dependency>
   
       
       <dependency>
          <groupId>com.fasterxml.jackson.coregroupId>
          <artifactId>jackson-annotationsartifactId>
          <version>2.13.4version>
       dependency>
   

创建项目

1. 创建普通java项目，创建main方法。

public static void main(String[] args) {
   ObjectMapper mapper = defaultObjectMapper();
   // 主机地址,端口号
   Proxy proxy = new Proxy(Proxy.Type.HTTP, new InetSocketAddress("127.0.0.1", port));
   //        设置token(需自行申请),增加timeout字段，修改默认超时时间
   OkHttpClient client = defaultClient("your_token", Duration.ofSeconds(100))
                .newBuilder()
                .proxy(proxy)
                .build();
   Retrofit retrofit = defaultRetrofit(client, mapper);
   OpenAiApi api = retrofit.create(OpenAiApi.class);
   OpenAiService service = new OpenAiService(api);
        CompletionRequest completionRequest = CompletionRequest.builder()
      //   设置提示
                .prompt("Somebody once told me the world is gonna roll me")
//                把ada模型修改为davinci模型，davinci模型功能更强大
                .model("text-davinci-003")
                .echo(true)
//               增加maxTokens字段，默认tokens数不够用
                .maxTokens(1000)
                .build();
        System.out.println("--------");
        service.createCompletion(completionRequest).getChoices().forEach(System.out::println);
        System.out.println("--------");
    }

2. 运行，在控制台窗口中显示

   CompletionChoice(text=Somebody once told me the world is gonna roll me

    I ain't the sharpest tool in the shed

    She was looking kind of dumb with her finger and her thumb
    In the shape of an "L" on her forehead,index=0, logprobs=null, finish_reason=stop)

注意(下方地址都需要魔法上网)：

1. 需要CompletionRequest请求参数设置maxTokens()字段，默认token数只能显示一部分返回文本(token:英文字母≈1:3~4,汉字:token≈2:1)。
   文本转token数地址,其他参数详情.
2. 须在调用的类OpenAiService里加上timeout参数即 Duration.ofSeconds(100)(Duration.ofSeconds( )的参数类型为long,单位为秒),否则可能会超时。

   OpenAiService service = new OpenAiService("your_token", Duration.ofSeconds(100));
   

3. 选择模型.text-davinci-003模型更好，想要了解其它模型可以访问https://platform.openai.com/docs/models/gpt-3。

修改

 问题：connect timed out
 解决方案：配置代理

1. 增加依赖包
2. 增加配置代码