一、什么是智谱AI
智谱AI(Zhipu AI)是一家致力于人工智能技术研发和应用的公司。该公司由清华大学背景的团队创立,专注于大模型技术的研究与推广。智谱AI在人工智能领域取得了显著成就,其发布的自研大模型GLM-4等产品。
二、SDK玩法
(一) 注册账号
进入官网(https://maas.aminer.cn/),注册账号实名后,将会赠送**有效期一个月**的体验包。
(二) 查看自己的API Key
注意:我们常见的API_KEY和API_SECRET,这里采用了统一为API key,使用 .这个符号进行划分。
举个栗子:yingzix688.xxxx。
那么,API_KEY:yingzix688
*** API_SECRET:xxxx***
大家只需要看自己的API key进行分割出来即可。
(三) 查阅官方github
1. 引入依赖
<dependency>
<groupId>cn.bigmodel.openapi</groupId>
<artifactId>oapi-java-sdk</artifactId>
<version>release-V4-2.0.0</version>
</dependency>
2. 官方示例代码
package com.zhipu.oapi.demo;
import com.alibaba.fastjson.JSON;
import com.fasterxml.jackson.annotation.JsonInclude;
import com.fasterxml.jackson.core.JsonProcessingException;
import com.fasterxml.jackson.core.type.TypeReference;
import com.fasterxml.jackson.databind.DeserializationFeature;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.fasterxml.jackson.databind.PropertyNamingStrategy;
import com.zhipu.oapi.ClientV4;
import com.zhipu.oapi.Constants;
import com.zhipu.oapi.service.v4.embedding.EmbeddingApiResponse;
import com.zhipu.oapi.service.v4.embedding.EmbeddingRequest;
import com.zhipu.oapi.service.v4.file.FileApiResponse;
import com.zhipu.oapi.service.v4.file.QueryFileApiResponse;
import com.zhipu.oapi.service.v4.file.QueryFilesRequest;
import com.zhipu.oapi.service.v4.fine_turning.*;
import com.zhipu.oapi.service.v4.image.CreateImageRequest;
import com.zhipu.oapi.service.v4.image.ImageApiResponse;
import com.zhipu.oapi.service.v4.model.*;
import io.reactivex.Flowable;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.concurrent.atomic.AtomicBoolean;
public class V4OkHttpClientTest {
private static final String API_KEY = "";
private static final String API_SECRET = "";
private static final ClientV4 client = new ClientV4.Builder(API_KEY,API_SECRET).build();
private static final ObjectMapper mapper = defaultObjectMapper();
public static ObjectMapper defaultObjectMapper() {
ObjectMapper mapper = new ObjectMapper();
mapper.configure(DeserializationFeature.FAIL_ON_UNKNOWN_PROPERTIES, false);
mapper.setSerializationInclusion(JsonInclude.Include.NON_NULL);
mapper.setPropertyNamingStrategy(PropertyNamingStrategy.SNAKE_CASE);
mapper.addMixIn(ChatFunction.class, ChatFunctionMixIn.class);
mapper.addMixIn(ChatCompletionRequest.class, ChatCompletionRequestMixIn.class);
mapper.addMixIn(ChatFunctionCall.class, ChatFunctionCallMixIn.class);
return mapper;
}
// 请自定义自己的业务id
private static final String requestIdTemplate = "mycompany-%d";
public static void main(String[] args) throws Exception {
System.setProperty("org.slf4j.simpleLogger.logFile", "System.out");
// 1. sse-invoke调用模型,使用标准Listener,直接返回结果
testSseInvoke();
// 2. invoke调用模型,直接返回结果
// testInvoke();
// 3. 异步调用
// String taskId = testAsyncInvoke();
// 4.异步查询
// testQueryResult(taskId);
// 5.文生图
// testCreateImage();
// 6. 图生文
// testImageToWord();
// 7. 向量模型
// testEmbeddings();
// 8.微调-上传微调数据集
// testUploadFile();
// 9.微调-查询上传文件列表
// testQueryUploadFileList();
// 10.微调-创建微调任务
// testCreateFineTuningJob();
// 11.微调-查询微调任务事件
// testQueryFineTuningJobsEvents();
// 12.微调-查询微调任务
// testRetrieveFineTuningJobs();
// 13.微调-查询个人微调任务
// testQueryPersonalFineTuningJobs();
// 14.微调-调用微调模型(参考模型调用接口,并替换成要调用模型的编码model)
}
private static void testQueryPersonalFineTuningJobs() {
QueryPersonalFineTuningJobRequest queryPersonalFineTuningJobRequest = new QueryPersonalFineTuningJobRequest();
queryPersonalFineTuningJobRequest.setLimit(1);
QueryPersonalFineTuningJobApiResponse queryPersonalFineTuningJobApiResponse = client.queryPersonalFineTuningJobs(queryPersonalFineTuningJobRequest);
System.out.println("model output:" + JSON.toJSONString(queryPersonalFineTuningJobApiResponse));
}
private static void testQueryFineTuningJobsEvents() {
QueryFineTuningJobRequest queryFineTuningJobRequest = new QueryFineTuningJobRequest();
queryFineTuningJobRequest.setJobId("ftjob-20240119114544390-zkgjb");
// queryFineTuningJobRequest.setLimit(1);
// queryFineTuningJobRequest.setAfter("1");
QueryFineTuningEventApiResponse queryFineTuningEventApiResponse = client.queryFineTuningJobsEvents(queryFineTuningJobRequest);
System.out.println("model output:" + JSON.toJSONString(queryFineTuningEventApiResponse));
}
/**
* 查询微调任务
*/
private static void testRetrieveFineTuningJobs() {
QueryFineTuningJobRequest queryFineTuningJobRequest = new QueryFineTuningJobRequest();
queryFineTuningJobRequest.setJobId("ftjob-20240119114544390-zkgjb");
// queryFineTuningJobRequest.setLimit(1);
// queryFineTuningJobRequest.setAfter("1");
QueryFineTuningJobApiResponse queryFineTuningJobApiResponse = client.retrieveFineTuningJobs(queryFineTuningJobRequest);
System.out.println("model output:" + JSON.toJSONString(queryFineTuningJobApiResponse));
}
/**
* 创建微调任务
*/
private static void testCreateFineTuningJob() {
FineTuningJobRequest request = new FineTuningJobRequest();
String requestId = String.format(requestIdTemplate, System.currentTimeMillis());
request.setRequestId(requestId);
request.setModel("chatglm3-6b");
request.setTraining_file("file-20240118082608327-kp8qr");
CreateFineTuningJobApiResponse createFineTuningJobApiResponse = client.createFineTuningJob(request);
System.out.println("model output:" + JSON.toJSONString(createFineTuningJobApiResponse));
}
/**
* 微调文件上传列表查询
*/
private static void testQueryUploadFileList() {
QueryFilesRequest queryFilesRequest = new QueryFilesRequest();
QueryFileApiResponse queryFileApiResponse = client.queryFilesApi(queryFilesRequest);
System.out.println("model output:" + JSON.toJSONString(queryFileApiResponse));
}
/**
* 微调上传数据集
*/
private static void testUploadFile() {
String filePath = "/Users/wujianguo/Downloads/transaction-data.jsonl";
String purpose = "fine-tune";
FileApiResponse fileApiResponse = client.invokeUploadFileApi(purpose, filePath);
System.out.println("model output:" + JSON.toJSONString(fileApiResponse));
}
private static void testEmbeddings() {
EmbeddingRequest embeddingRequest = new EmbeddingRequest();
embeddingRequest.setInput("hello world");
embeddingRequest.setModel(Constants.ModelEmbedding2);
EmbeddingApiResponse apiResponse = client.invokeEmbeddingsApi(embeddingRequest);
System.out.println("model output:" + JSON.toJSONString(apiResponse));
}
/**
* 图生文
*/
private static void testImageToWord() {
List<ChatMessage> messages = new ArrayList<>();
List<Map<String, Object>> contentList = new ArrayList<>();
Map<String, Object> textMap = new HashMap<>();
textMap.put("type", "text");
textMap.put("text", "图里有什么");
Map<String, Object> typeMap = new HashMap<>();
typeMap.put("type", "image_url");
Map<String, Object> urlMap = new HashMap<>();
urlMap.put("url", "https://cdn.bigmodel.cn/enterpriseAc/3f328152-e15c-420c-803d-6684a9f551df.jpeg?attname=24.jpeg");
typeMap.put("image_url", urlMap);
contentList.add(textMap);
contentList.add(typeMap);
ChatMessage chatMessage = new ChatMessage(ChatMessageRole.USER.value(), contentList);
messages.add(chatMessage);
String requestId = String.format(requestIdTemplate, System.currentTimeMillis());
ChatCompletionRequest chatCompletionRequest = ChatCompletionRequest.builder()
.model(Constants.ModelChatGLM4V)
.stream(Boolean.FALSE)
.invokeMethod(Constants.invokeMethod)
.messages(messages)
.requestId(requestId)
.build();
ModelApiResponse modelApiResponse = client.invokeModelApi(chatCompletionRequest);
System.out.println("model output:" + JSON.toJSONString(modelApiResponse));
}
private static void testCreateImage() {
CreateImageRequest createImageRequest = new CreateImageRequest();
createImageRequest.setModel(Constants.ModelCogView);
// createImageRequest.setPrompt("画一个温顺可爱的小狗");
ImageApiResponse imageApiResponse = client.createImage(createImageRequest);
System.out.println("imageApiResponse:" + JSON.toJSONString(imageApiResponse));
}
/**
* sse调用
*/
private static void testSseInvoke() {
List<ChatMessage> messages = new ArrayList<>();
ChatMessage chatMessage = new ChatMessage(ChatMessageRole.USER.value(), "ChatGLM和你哪个更强大");
// ChatMessage chatMessage = new ChatMessage(ChatMessageRole.USER.value(), "你能帮我查询2024年1月1日从北京南站到上海的火车票吗?");
messages.add(chatMessage);
String requestId = String.format(requestIdTemplate, System.currentTimeMillis());
// 函数调用参数构建部分
List<ChatTool> chatToolList = new ArrayList<>();
ChatTool chatTool = new ChatTool();
chatTool.setType(ChatToolType.FUNCTION.value());
ChatFunctionParameters chatFunctionParameters = new ChatFunctionParameters();
chatFunctionParameters.setType("object");
Map<String, Object> properties = new HashMap<>();
properties.put("departure", new HashMap<String, Object>() {{
put("type", "string");
put("description", "出发城市或车站");
}});
properties.put("destination", new HashMap<String, Object>() {{
put("type", "string");
put("description", "目的地城市或车站");
}});
properties.put("date", new HashMap<String, Object>() {{
put("type", "string");
put("description", "要查询的车次日期");
}});
List<String> required = new ArrayList<>();
required.add("departure");
required.add("destination");
required.add("date");
chatFunctionParameters.setProperties(properties);
ChatFunction chatFunction = ChatFunction.builder()
.name("query_train_info")
.description("根据用户提供的信息,查询对应的车次")
.parameters(chatFunctionParameters)
.required(required)
.build();
chatTool.setFunction(chatFunction);
chatToolList.add(chatTool);
ChatCompletionRequest chatCompletionRequest = ChatCompletionRequest.builder()
.model(Constants.ModelChatGLM4)
.stream(Boolean.TRUE)
.messages(messages)
.requestId(requestId)
.tools(chatToolList)
.toolChoice("auto")
.build();
ModelApiResponse sseModelApiResp = client.invokeModelApi(chatCompletionRequest);
if (sseModelApiResp.isSuccess()) {
AtomicBoolean isFirst = new AtomicBoolean(true);
ChatMessageAccumulator chatMessageAccumulator = mapStreamToAccumulator(sseModelApiResp.getFlowable())
.doOnNext(accumulator -> {
{
if (isFirst.getAndSet(false)) {
System.out.print("Response: ");
}
if (accumulator.getDelta() != null && accumulator.getDelta().getTool_calls() != null) {
String jsonString = mapper.writeValueAsString(accumulator.getDelta().getTool_calls());
System.out.println("tool_calls: " + jsonString);
}
if (accumulator.getDelta() != null && accumulator.getDelta().getContent() != null) {
System.out.print(accumulator.getDelta().getContent());
}
}
})
.doOnComplete(System.out::println)
.lastElement()
.blockingGet();
Choice choice = new Choice(chatMessageAccumulator.getChoice().getFinishReason(), 0L, chatMessageAccumulator.getDelta());
List<Choice> choices = new ArrayList<>();
choices.add(choice);
ModelData data = new ModelData();
data.setChoices(choices);
data.setUsage(chatMessageAccumulator.getUsage());
data.setId(chatMessageAccumulator.getId());
data.setCreated(chatMessageAccumulator.getCreated());
data.setRequestId(chatCompletionRequest.getRequestId());
sseModelApiResp.setFlowable(null);
sseModelApiResp.setData(data);
}
System.out.println("model output:" + JSON.toJSONString(sseModelApiResp));
}
public static Flowable<ChatMessageAccumulator> mapStreamToAccumulator(Flowable<ModelData> flowable) {
return flowable.map(chunk -> {
return new ChatMessageAccumulator(chunk.getChoices().get(0).getDelta(), null, chunk.getChoices().get(0), chunk.getUsage(), chunk.getCreated(), chunk.getId());
});
}
/**
* 同步调用
*/
private static void testInvoke() {
List<ChatMessage> messages = new ArrayList<>();
ChatMessage chatMessage = new ChatMessage(ChatMessageRole.USER.value(), "ChatGLM和你哪个更强大");
messages.add(chatMessage);
String requestId = String.format(requestIdTemplate, System.currentTimeMillis());
// 函数调用参数构建部分
List<ChatTool> chatToolList = new ArrayList<>();
ChatTool chatTool = new ChatTool();
chatTool.setType(ChatToolType.FUNCTION.value());
ChatFunctionParameters chatFunctionParameters = new ChatFunctionParameters();
chatFunctionParameters.setType("object");
Map<String, Object> properties = new HashMap<>();
properties.put("location", new HashMap<String, Object>() {{
put("type", "string");
put("description", "城市,如:北京");
}});
properties.put("unit", new HashMap<String, Object>() {{
put("type", "string");
put("enum", new ArrayList<String>() {{
add("celsius");
add("fahrenheit");
}});
}});
chatFunctionParameters.setProperties(properties);
ChatFunction chatFunction = ChatFunction.builder()
.name("get_weather")
.description("Get the current weather of a location")
.parameters(chatFunctionParameters)
.build();
chatTool.setFunction(chatFunction);
chatToolList.add(chatTool);
ChatCompletionRequest chatCompletionRequest = ChatCompletionRequest.builder()
.model(Constants.ModelChatGLM4)
.stream(Boolean.FALSE)
.invokeMethod(Constants.invokeMethod)
.messages(messages)
.requestId(requestId)
.tools(chatToolList)
.toolChoice("auto")
.build();
ModelApiResponse invokeModelApiResp = client.invokeModelApi(chatCompletionRequest);
try {
System.out.println("model output:" + mapper.writeValueAsString(invokeModelApiResp));
} catch (JsonProcessingException e) {
e.printStackTrace();
}
}
/**
* 异步调用
*/
private static String testAsyncInvoke() {
List<ChatMessage> messages = new ArrayList<>();
ChatMessage chatMessage = new ChatMessage(ChatMessageRole.USER.value(), "ChatLM和你哪个更强大");
messages.add(chatMessage);
String requestId = String.format(requestIdTemplate, System.currentTimeMillis());
// 函数调用参数构建部分
List<ChatTool> chatToolList = new ArrayList<>();
ChatTool chatTool = new ChatTool();
chatTool.setType(ChatToolType.FUNCTION.value());
ChatFunctionParameters chatFunctionParameters = new ChatFunctionParameters();
chatFunctionParameters.setType("object");
Map<String, Object> properties = new HashMap<>();
properties.put("location", new HashMap<String, Object>() {{
put("type", "string");
put("description", "城市,如:北京");
}});
properties.put("unit", new HashMap<String, Object>() {{
put("type", "string");
put("enum", new ArrayList<String>() {{
add("celsius");
add("fahrenheit");
}});
}});
chatFunctionParameters.setProperties(properties);
ChatFunction chatFunction = ChatFunction.builder()
.name("get_weather")
.description("Get the current weather of a location")
.parameters(chatFunctionParameters)
.build();
chatTool.setFunction(chatFunction);
chatToolList.add(chatTool);
ChatCompletionRequest chatCompletionRequest = ChatCompletionRequest.builder()
.model(Constants.ModelChatGLM4)
.stream(Boolean.FALSE)
.invokeMethod(Constants.invokeMethodAsync)
.messages(messages)
.requestId(requestId)
.tools(chatToolList)
.toolChoice("auto")
.build();
ModelApiResponse invokeModelApiResp = client.invokeModelApi(chatCompletionRequest);
System.out.println("model output:" + JSON.toJSONString(invokeModelApiResp));
return invokeModelApiResp.getData().getTaskId();
}
/**
* 查询异步结果
*
* @param taskId
*/
private static void testQueryResult(String taskId) {
QueryModelResultRequest request = new QueryModelResultRequest();
request.setTaskId(taskId);
QueryModelResultResponse queryResultResp = client.queryModelResult(request);
try {
System.out.println("model output:" + mapper.writeValueAsString(queryResultResp));
} catch (JsonProcessingException e) {
e.printStackTrace();
}
}
}
A. 填充自己的信息
B. 启动示例
这里解释一下四个的区别
- testSseInvoke 使用的是逐渐输出,AI回答的结果是一段一段的展示。
- testInvoke 使用的是同步执行,当AI全部的回答都输出后才会展示出来。
- testAsyncInvoke 与testQueryResult 搭配使用,先通过testAsyncInvoke 让AI去执行,直接返回一个成功或者失败,之后通过获得的taskId,再用testQueryResult去查询获得结果即可。这个过程实践过Bi项目的小伙伴应该深有体会。
C. 结果展示
3.参数阅读
这里补充下,如果你要修改问题,只需要修改content参数中的值即可。相当于问AI问题。
其实最关键的就是前三个
A. model
你要选择哪个模型, 例如选择GLM-4 还是GLM-3-Turbo
B. messages
这里需要考虑两个值,一个是role,一般为user即可。role的值官方已经给我们枚举了,只需要调用即可。
剩下的则是我们需要自己填入的content
C. request_id
这个是区分我们每次上传的任务,保证唯一性,可以自己上传一个类似于雪花算法的ID,用户端不传的话平台也会自动生成。
其他参数目前影子测试完前五个方法后发现使用官方默认的即可。只需要你调整好代码的位置以及content的值即可。
剩下的参数,如果你需要使用微调或者向量知识库等高阶玩法,则根据官方文档调整即可。很多地方已经自带了枚举值,只需要直接选择填充。
最后,大家可以用这一个月的免费额度,打造一个自己的AI小工具使用,更多玩法,由大家一起探索。
我是程序员影子,一名以Java为主,其余时间探索AI+编程的程序猿。
以上就是本次分享的所有内容,感兴趣的朋友点个关注呀,感谢大家啦~
版权归原作者 程序员影子 所有, 如有侵权,请联系我们删除。