[KYUUBI #4810] [CHAT] Request and use a sginle choice for chat completion

### _Why are the changes needed?_

- explicitly set `n` to 1 in ChatGPT chat completion request (default to 1, https://platform.openai.com/docs/api-reference/chat/create#chat/create-n)
- use the only one choice of the chat completion response

### _How was this patch tested?_
- [ ] Add some test cases that check the changes thoroughly including negative and positive cases if possible

- [ ] Add screenshots for manual tests if appropriate

- [x] [Run test](https://kyuubi.readthedocs.io/en/master/develop_tools/testing.html#running-tests) locally before make a pull request

Closes #4810 from bowenliang123/chat-onechoice.

Closes #4810

f221de4e8 [bowenliang] one message

Authored-by: bowenliang <bowenliang@apache.org>
Signed-off-by: bowenliang <bowenliang@apache.org>

externals/kyuubi-chat-engine/src/main/scala/org/apache/kyuubi/engine/chat/provider/ChatGPTProvider.scala

@@ -83,9 +83,10 @@ class ChatGPTProvider(conf: KyuubiConf) extends ChatProvider {
         .model(conf.get(KyuubiConf.ENGINE_CHAT_GPT_MODEL))
         .messages(messages.asScala.toList.asJava)
         .user(sessionUser.orNull)
+        .n(1)
         .build()
-      val responseText = openAiService.createChatCompletion(completionRequest).getChoices.asScala
-        .map(c => c.getMessage.getContent).mkString
+      val responseText = openAiService.createChatCompletion(completionRequest)
+        .getChoices.get(0).getMessage.getContent
       responseText
     } catch {
       case e: Throwable =>