溫馨提示×

溫馨提示×

您好,登錄后才能下訂單哦!

密碼登錄×
登錄注冊×
其他方式登錄
點擊 登錄注冊 即表示同意《億速云用戶服務(wù)條款》

java如何調(diào)用chatgpt接口來實現(xiàn)專屬于自己的人工智能助手

發(fā)布時間:2023-03-27 10:48:03 來源:億速云 閱讀:121 作者:iii 欄目:開發(fā)技術(shù)

這篇文章主要介紹“java如何調(diào)用chatgpt接口來實現(xiàn)專屬于自己的人工智能助手”的相關(guān)知識,小編通過實際案例向大家展示操作過程,操作方法簡單快捷,實用性強(qiáng),希望這篇“java如何調(diào)用chatgpt接口來實現(xiàn)專屬于自己的人工智能助手”文章能幫助大家解決問題。

導(dǎo)包

對于http調(diào)用,我使用的是比較流行的httpclient5,然后直接創(chuàng)建了一個springboot項目,方便以后對外提供接口。

    <parent>
        <artifactId>spring-boot-starter-parent</artifactId>
        <groupId>org.springframework.boot</groupId>
        <version>2.5.3</version>
    </parent>

    <dependencies>

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

        <dependency>
            <groupId>org.projectlombok</groupId>
            <artifactId>lombok</artifactId>
        </dependency>

        <!-- https://mvnrepository.com/artifact/org.apache.httpcomponents.client5/httpclient5 -->
        <dependency>
            <groupId>org.apache.httpcomponents.client5</groupId>
            <artifactId>httpclient5</artifactId>
            <version>5.2.1</version>
        </dependency>

    </dependencies>

基本說明

在編寫代碼之前,這個先給出HttpClient的Api文檔 api文檔

java如何調(diào)用chatgpt接口來實現(xiàn)專屬于自己的人工智能助手

我們在編寫代碼之前需要了解官方提供的接口如何進(jìn)行訪問以及返回的結(jié)果是什么

請求參數(shù)

官方文檔地址為文檔,請求參數(shù)必須填寫的內(nèi)容如下

{
  "model": "gpt-3.5-turbo",
  "messages": [{"role": "user", "content": "Hello!"}]
}

一個是model,一個是messages。model根據(jù)自己的情況來選擇,聊天的話就是gpt-3.5-turbo,下面的messages里面包含n個對象,每個對象有role和content,role表示角色,content表示內(nèi)容。
下面為官方文檔中的解釋

java如何調(diào)用chatgpt接口來實現(xiàn)專屬于自己的人工智能助手

簡單理解就是我們要問問題,role就是user。如果要實現(xiàn)連續(xù)對話,那么就將返回的返回內(nèi)容設(shè)置到messages中,role設(shè)置為返回的role。

響應(yīng)參數(shù)

下面直接給出響應(yīng)的內(nèi)容

{
 'id': 'chatcmpl-6p9XYPYSTTRi0xEviKjjilqrWU2Ve',
 'object': 'chat.completion',
 'created': 1677649420,
 'model': 'gpt-3.5-turbo',
 'usage': {'prompt_tokens': 56, 'completion_tokens': 31, 'total_tokens': 87},
 'choices': [
   {
    'message': {
      'role': 'assistant',
      'content': 'The 2020 World Series was played in Arlington, Texas at the Globe Life Field, which was the new home stadium for the Texas Rangers.'},
    'finish_reason': 'stop',
    'index': 0
   }
  ]
}

我們問問題的答案就在choices.message下的content中,而role就代表了chatGpt扮演的角色??吹竭@我們就應(yīng)該知道該干嘛了吧肯定是創(chuàng)建對應(yīng)的VO類啊。

創(chuàng)建請求和響應(yīng)的VO類

下面5個類就對應(yīng)了我們發(fā)送和接收的各種信息

ChatGptMessage類

@Data
@NoArgsConstructor
@AllArgsConstructor
public class ChatGptMessage {
        String role;
        String content;
}

ChatGptRequestParameter 類

@Data
@NoArgsConstructor
@AllArgsConstructor
public class ChatGptRequestParameter {

    String model = "gpt-3.5-turbo";

    List<ChatGptMessage> messages = new ArrayList<>();

    public void addMessages(ChatGptMessage message) {
        this.messages.add(message);
    }

}

ChatGptResponseParameter 類

@Data
@NoArgsConstructor
@AllArgsConstructor
public class ChatGptResponseParameter {

    String id;
    String object;
    String created;
    String model;
    Usage usage;
    List<Choices> choices;
}

Choices 類

@Data
@NoArgsConstructor
@AllArgsConstructor
public class Choices {

    ChatGptMessage message;
    String finish_reason;
    Integer index;
}

Usage 類

@Data
@NoArgsConstructor
@AllArgsConstructor
public class Usage {

    String prompt_tokens;
    String completion_tokens;
    String total_tokens;
}

代碼編寫

不說廢話,首先創(chuàng)建一個CustomChatGpt類

public class CustomChatGpt {

}

然后定義一些成員屬性

    /**
     * 自己chatGpt的ApiKey
     */
    private String apiKey;
    /**
     * 使用的模型
     */
    private String model = "gpt-3.5-turbo-0301";
    /**
     * 對應(yīng)的請求接口
     */
    private String url = "https://api.openai.com/v1/chat/completions";
    /**
     * 默認(rèn)編碼
     */
    private Charset charset = StandardCharsets.UTF_8;
        /**
     * 創(chuàng)建一個ChatGptRequestParameter,用于攜帶請求參數(shù)
     */
    private ChatGptRequestParameter chatGptRequestParameter = new ChatGptRequestParameter();

提供一個ApiKey的構(gòu)造器,創(chuàng)建該對象必須要傳入ApiKey

    public CustomChatGpt(String apiKey) {
        this.apiKey = apiKey;
    }

定義一個響應(yīng)超時時間

    /**
     * 響應(yīng)超時時間,毫秒
     */
    private int responseTimeout = 10000;

    public void setResponseTimeout(int responseTimeout) {
        this.responseTimeout = responseTimeout;
    }

編寫一個getAnswer方法,要求傳入一個CloseableHttpClient和一個問題

    public String getAnswer(CloseableHttpClient client, String question) {
    
    }

繼續(xù)實現(xiàn)方法,下面會完成一些參數(shù)的創(chuàng)建和設(shè)置

        // 創(chuàng)建一個HttpPost
        HttpPost httpPost = new HttpPost(url);
        // 創(chuàng)建一個ObjectMapper,用于解析和創(chuàng)建json
        ObjectMapper objectMapper = new ObjectMapper();
        // 設(shè)置請求參數(shù)
        chatGptRequestParameter.addMessages(new ChatGptMessage("user", question));
        HttpEntity httpEntity = null;
        try {
            // 對象轉(zhuǎn)換為json字符串
            httpEntity = new StringEntity(objectMapper.writeValueAsString(chatGptRequestParameter), charset);
        } catch (JsonProcessingException e) {
            System.out.println(question + "->json轉(zhuǎn)換異常");
            return null;
        }
        httpPost.setEntity(httpEntity);

下面會完成一些配置的設(shè)置

        // 設(shè)置請求頭
        httpPost.setHeader(HttpHeaders.CONTENT_TYPE, "application/json");
        // 設(shè)置登錄憑證
        httpPost.setHeader(HttpHeaders.AUTHORIZATION, "Bearer " + apiKey);

        // 用于設(shè)置超時時間
        RequestConfig config = RequestConfig
                .custom()
                .setResponseTimeout(responseTimeout, TimeUnit.MILLISECONDS)
                .build();
        httpPost.setConfig(config);

下面代碼會提交請求,解析響應(yīng),最后返回對應(yīng)問題的答案

        try {
            // 提交請求
            return client.execute(httpPost, response -> {
                // 得到返回的內(nèi)容
                String resStr = EntityUtils.toString(response.getEntity(), charset);
                // 轉(zhuǎn)換為對象
                ChatGptResponseParameter responseParameter = objectMapper.readValue(resStr, ChatGptResponseParameter.class);
                String ans = "";
                // 遍歷所有的Choices(一般都只有一個)
                for (Choices choice : responseParameter.getChoices()) {
                    ChatGptMessage message = choice.getMessage();
                    chatGptRequestParameter.addMessages(new ChatGptMessage(message.getRole(), message.getContent()));
                    String s = message.getContent().replaceAll("\n+", "\n");
                    ans += s;
                }
                // 返回信息
                return ans;
            });
        } catch (IOException e) {
            e.printStackTrace();
        }
        // 發(fā)生異常,移除剛剛添加的ChatGptMessage
        chatGptRequestParameter.getMessages().remove(chatGptRequestParameter.getMessages().size()-1);
        return "您當(dāng)前的網(wǎng)絡(luò)無法訪問";

下面給出這個類的完整代碼

import com.fasterxml.jackson.core.JsonProcessingException;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.ttpfx.vo.ChatGptMessage;
import com.ttpfx.vo.ChatGptRequestParameter;
import com.ttpfx.vo.ChatGptResponseParameter;
import com.ttpfx.vo.Choices;
import org.apache.hc.client5.http.classic.methods.HttpPost;
import org.apache.hc.client5.http.config.RequestConfig;
import org.apache.hc.client5.http.impl.classic.CloseableHttpClient;
import org.apache.hc.core5.http.HttpEntity;
import org.apache.hc.core5.http.HttpHeaders;
import org.apache.hc.core5.http.io.entity.EntityUtils;
import org.apache.hc.core5.http.io.entity.StringEntity;

import java.io.IOException;
import java.nio.charset.Charset;
import java.nio.charset.StandardCharsets;
import java.util.concurrent.TimeUnit;

/**
 * @author ttpfx
 * @date 2023/3/23
 */
public class CustomChatGpt {
    /**
     * 自己chatGpt的ApiKey
     */
    private String apiKey;
    /**
     * 使用的模型
     */
    private String model = "gpt-3.5-turbo-0301";
    /**
     * 對應(yīng)的請求接口
     */
    private String url = "https://api.openai.com/v1/chat/completions";
    /**
     * 默認(rèn)編碼
     */
    private Charset charset = StandardCharsets.UTF_8;


    /**
     * 創(chuàng)建一個ChatGptRequestParameter,用于攜帶請求參數(shù)
     */
    private ChatGptRequestParameter chatGptRequestParameter = new ChatGptRequestParameter();

    /**
     * 相應(yīng)超時時間,毫秒
     */
    private int responseTimeout = 1000;

    public void setResponseTimeout(int responseTimeout) {
        this.responseTimeout = responseTimeout;
    }

    public CustomChatGpt(String apiKey) {
        this.apiKey = apiKey;
    }

    public String getAnswer(CloseableHttpClient client, String question) {

        // 創(chuàng)建一個HttpPost
        HttpPost httpPost = new HttpPost(url);
        // 創(chuàng)建一個ObjectMapper,用于解析和創(chuàng)建json
        ObjectMapper objectMapper = new ObjectMapper();

        // 設(shè)置請求參數(shù)
        chatGptRequestParameter.addMessages(new ChatGptMessage("user", question));
        HttpEntity httpEntity = null;
        try {
            // 對象轉(zhuǎn)換為json字符串
            httpEntity = new StringEntity(objectMapper.writeValueAsString(chatGptRequestParameter), charset);
        } catch (JsonProcessingException e) {
            System.out.println(question + "->json轉(zhuǎn)換異常");
            return null;
        }
        httpPost.setEntity(httpEntity);


        // 設(shè)置請求頭
        httpPost.setHeader(HttpHeaders.CONTENT_TYPE, "application/json");
        // 設(shè)置登錄憑證
        httpPost.setHeader(HttpHeaders.AUTHORIZATION, "Bearer " + apiKey);

        // 用于設(shè)置超時時間
        RequestConfig config = RequestConfig
                .custom()
                .setResponseTimeout(responseTimeout, TimeUnit.MILLISECONDS)
                .build();
        httpPost.setConfig(config);
        try {
            // 提交請求
            return client.execute(httpPost, response -> {
                // 得到返回的內(nèi)容
                String resStr = EntityUtils.toString(response.getEntity(), charset);
                // 轉(zhuǎn)換為對象
                ChatGptResponseParameter responseParameter = objectMapper.readValue(resStr, ChatGptResponseParameter.class);
                String ans = "";
                // 遍歷所有的Choices(一般都只有一個)
                for (Choices choice : responseParameter.getChoices()) {
                    ChatGptMessage message = choice.getMessage();
                    chatGptRequestParameter.addMessages(new ChatGptMessage(message.getRole(), message.getContent()));
                    String s = message.getContent().replaceAll("\n+", "\n");
                    ans += s;
                }
                // 返回信息
                return ans;
            });
        } catch (IOException e) {
            e.printStackTrace();
        }
        // 發(fā)生異常,移除剛剛添加的ChatGptMessage
        chatGptRequestParameter.getMessages().remove(chatGptRequestParameter.getMessages().size()-1);
        return "您當(dāng)前的網(wǎng)絡(luò)無法訪問";
    }
}

使用

下面就是測試代碼,我們只需要傳入一個CloseableHttpClient 和 question 即可

public class Test {
    public static void main(String[] args) throws IOException {
        CloseableHttpClient httpClient = HttpClients.createDefault();
        String apiKey = "自己的ApiKey";
        CustomChatGpt customChatGpt = new CustomChatGpt(apiKey);
        // 根據(jù)自己的網(wǎng)絡(luò)設(shè)置吧
        customChatGpt.setResponseTimeout(20000);
        while (true) {
            System.out.print("\n請輸入問題(q退出):");
            String question = new Scanner(System.in).nextLine();
            if ("q".equals(question)) break;
            long start = System.currentTimeMillis();
            String answer = customChatGpt.getAnswer(httpClient, question);
            long end = System.currentTimeMillis();
            System.out.println("該回答花費時間為:" + (end - start) / 1000.0 + "秒");
            System.out.println(answer);
        }
        httpClient.close();
    }
}

最后說明

對于ApiKey,只能說難者不會,會者不難,這個沒辦法教。

如果代碼無法運行,或者運行速度及其緩慢,請使用代理,在HttpClient里面可以很輕松的使用代理

        String proxyIp = "127.0.0.1";
        int proxyPort = 7890;
        HttpHost httpHost = new HttpHost(proxyIp, proxyPort);

上面就是一個示例,對于代理,這里也就無法繼續(xù)進(jìn)行說明了。

如果我們完成了上面的功能,是不是就能夠?qū)ν馓峁┙涌?,然后寫一個自己的網(wǎng)頁端的ChatGpt或者弄一個聊天機(jī)器人呢?當(dāng)然沒問題啊

關(guān)于“java如何調(diào)用chatgpt接口來實現(xiàn)專屬于自己的人工智能助手”的內(nèi)容就介紹到這里了,感謝大家的閱讀。如果想了解更多行業(yè)相關(guān)的知識,可以關(guān)注億速云行業(yè)資訊頻道,小編每天都會為大家更新不同的知識點。

向AI問一下細(xì)節(jié)

免責(zé)聲明:本站發(fā)布的內(nèi)容(圖片、視頻和文字)以原創(chuàng)、轉(zhuǎn)載和分享為主,文章觀點不代表本網(wǎng)站立場,如果涉及侵權(quán)請聯(lián)系站長郵箱:is@yisu.com進(jìn)行舉報,并提供相關(guān)證據(jù),一經(jīng)查實,將立刻刪除涉嫌侵權(quán)內(nèi)容。

AI