grt1228 / chatgpt-steam-output Goto Github PK

java.lang.ClassCastException: org.springframework.web.servlet.mvc.method.annotation.SseEmitter cannot be cast to java.lang.String
at com.unfbx.chatgptsteamoutput.websocket.WebSocketServer.onMessage(WebSocketServer.java:120)

是不是还是把用户的问题和ai的回答再发给ai

被墙是怎么解决的

2024-02-23 16:50:12.129 INFO 5164 --- [.openai.com/...] u.c.l.OpenAIWebSocketEventSourceListener : OpenAI返回数据：{"id":"chatcmpl-8vLXvRaXtL18nmF0bMkm04uHZRlk0","object":"chat.completion.chunk","created":1708678211,"model":"gpt-3.5-turbo-0125","system_fingerprint":"fp_86156a94a0","choices":[{"index":0,"delta":{"role":"assistant","content":""},"logprobs":null,"finish_reason":null}]}
2024-02-23 16:50:12.149 ERROR 5164 --- [.openai.com/...] u.c.l.OpenAIWebSocketEventSourceListener : OpenAI sse连接异常data：，异常：{}

com.fasterxml.jackson.databind.exc.UnrecognizedPropertyException: Unrecognized field "system_fingerprint" (class com.unfbx.chatgpt.entity.chat.ChatCompletionResponse), not marked as ignorable (6 known properties: "usage", "choices", "created", "model", "id", "object"])
at [Source: (String)"{"id":"chatcmpl-8vLXvRaXtL18nmF0bMkm04uHZRlk0","object":"chat.completion.chunk","created":1708678211,"model":"gpt-3.5-turbo-0125","system_fingerprint":"fp_86156a94a0","choices":[{"index":0,"delta":{"role":"assistant","content":""},"logprobs":null,"finish_reason":null}]}"; line: 1, column: 153] (through reference chain: com.unfbx.chatgpt.entity.chat.ChatCompletionResponse["system_fingerprint"])

把onEvent里的回答拼接成一个字符串，用TikTokensUtil.tokens得出的记过比OpenAISSEEventSourceListener里累加的token要多10%，不太明白为什么要采用累加的方法计算token？

gpt3.5接入后，在聊天窗口问：你是那个版本，gpt回复:我是OpenAI GPT-3的AI语言模型，是目前最先进的自然语言处理模型之一；为什么会出现这个问题呢？

问题如下

idea的代理已经配置成功

能ping通网页版，但无法通过demo运行，望各位大佬解答，谢谢！

https://github.com/TheoKanning/openai-java 这个项目中的类似于 Flowable 的返回

1.使用环境
jdk8 linux服务器

1.0.11版本
在代码中已经设置了代理，并且sse是能够返回数据的
但是使用websocket就总是提示创建失败

但奇怪的是，偶尔情况下websocket又能够创建成功

存在内存泄漏：okhttpclient a connection to http was leaked

使用websocket案例后，无法返回答案

有这样的一个需求。APP请求java后端，java后端请求AI中台，AI中台再请求OPENAI的接口，答案进行流式返回。这个应该如何修改呢？

流式问答如何做上下文关联，要关联AI的回答？

2023-03-15T08:52:50.417+08:00 ERROR 25819 --- [io-10001-exec-1] o.a.coyote.http11.Http11NioProtocol : Failed to complete processing of a request
2594
2595 java.lang.OutOfMemoryError: Java heap space
2596 at java.base/java.nio.HeapByteBuffer.(HeapByteBuffer.java:64) ~[na:na]
2597 at java.base/java.nio.ByteBuffer.allocate(ByteBuffer.java:363) ~[na:na]
2598 at org.apache.coyote.http11.Http11InputBuffer.init(Http11InputBuffer.java:757) ~[tomcat-embed-core-10.1.5.jar!/:na]
2599 at org.apache.coyote.http11.Http11Processor.setSocketWrapper(Http11Processor.java:501) ~[tomcat-embed-core-10.1.5.jar!/:na]
2600 at org.apache.coyote.http11.Http11Processor.service(Http11Processor.java:258) ~[tomcat-embed-core-10.1.5.jar!/:na]
2601 at org.apache.coyote.AbstractProcessorLight.process(AbstractProcessorLight.java:65) ~[tomcat-embed-core-10.1.5.jar!/:na]
2602 at org.apache.coyote.AbstractProtocol$ConnectionHandler.process(AbstractProtocol.java:859) ~[tomcat-embed-core-10.1.5.jar!/:na]
2603 at org.apache.tomcat.util.net.NioEndpoint$SocketProcessor.doRun(NioEndpoint.java:1734) ~[tomcat-embed-core-10.1.5.jar!/:na]
2604 at org.apache.tomcat.util.net.SocketProcessorBase.run(SocketProcessorBase.java:52) ~[tomcat-embed-core-10.1.5.jar!/:na]
2605 at org.apache.tomcat.util.threads.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1191) ~[tomcat-embed-core-10.1.5.jar!/:na]
2606 at org.apache.tomcat.util.threads.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:659) ~[tomcat-embed-core-10.1.5.jar!/:na]
2607 at org.apache.tomcat.util.threads.TaskThread$WrappingRunnable.run(TaskThread.java:61) ~[tomcat-embed-core-10.1.5.jar!/:na]
2608 at java.base/java.lang.Thread.run(Thread.java:833) ~[na:na]
2609
2610 2023-03-15T08:52:50.422+08:00 ERROR 25819 --- [io-10001-exec-1] org.apache.tomcat.util.net.NioEndpoint : Error running socket processor

我已经设置了sse的超时时间为400，如果设置太短由于重连会重复输出，有啥好的建议吗
SseEmitter sseEmitter = new SseEmitter(400*1000l);

请问一直请求超时是怎么回事呀，本机有用clash设置代理，用postman请求是有返回的
2023-05-05 11:19:15.598 INFO 12776 --- [.openai.com/...] c.u.chatgpt.interceptor.OpenAILogger : OkHttp-------->:<-- HTTP FAILED: java.net.ConnectException: Failed to connect to api.openai.com/74.86.226.234:443

OkHttp-------->:<-- HTTP FAILED: javax.net.ssl.SSLHandshakeException: Received fatal alert: handshake_failure

之前一直运行没问题，这几天一直报错如上。
但是部署到服务器上就能正常运行，多台设备clone后运行都报同样问题，怪了。求解~（代码已经更新到git上同步最新的了还是会存在这个问题）

public SseEmitter chat(@RequestParam("message") String msg, @RequestHeader Map<String, String> headers) throws IOException {
SseEmitter sseEmitter = new SseEmitter(0L);

改成上面代码就好了

@slf4j
public class ConsoleEventSourceListener extends EventSourceListener {

@Override
public void onOpen(EventSource eventSource, Response response) {
    log.info("OpenAI建立sse连接...");
}

@Override
public void onEvent(EventSource eventSource, String id, String type, String data) {
    log.info("OpenAI返回数据：{}", data);
    if (data.equals("[DONE]")) {
        log.info("OpenAI返回数据结束了");
        return;
    }
}

@Override
public void onClosed(EventSource eventSource) {
    log.info("OpenAI关闭sse连接...");
}

@SneakyThrows
@Override
public void onFailure(EventSource eventSource, Throwable t, Response response) {
    if(Objects.isNull(response)){
        log.error("OpenAI  sse连接异常:{}", t);
        eventSource.cancel();
        return;
    }
    ResponseBody body = response.body();
    if (Objects.nonNull(body)) {
        log.error("OpenAI  sse连接异常data：{}，异常：{}", body.string(), t);
    } else {
        log.error("OpenAI  sse连接异常data：{}，异常：{}", response, t);
    }
    eventSource.cancel();
}

}

我现在跟前端对接有点问题，前端使用的http请求直接 responseBody.charStream() ，但是我返回的数据是一段一段的。谢谢大佬，我看你postman的测试好像是对的

server端只能支持打开5个连接，使用多台电脑，不同浏览器，请问是什么原因