What's the proper way to use ChatCompletion in stream mode?

[DEAN-Cherry]

According to the doc, it says you should use CreateCompletionAsStream to have streaming. However, there is a boolean property in ChatCompletionCreateRequest. I got confused that what is the proper way to use stream mode. There are 4 kinds of combination to use it, like CreateCompletion + Stream=true and CreateCompletionAsStream + Stream=ture

var completionResult = openAiService.ChatCompletion.CreateCompletionAsStream(
                new ChatCompletionCreateRequest()
                {
                    Messages = new List<ChatMessage>
                    {
                        ChatMessage.FromSystem("You are a helpful assistant."),
                        ChatMessage.FromUser("Some text here."),
                    },
                    // Stream = true,
                    Model = OpenAIModels.Gpt_4_1106_preview,
                    // MaxTokens = 50 //optional
                });

And I found that when uses CreateCompletionAsStream both Delta.Content and Message.Content are available. What's the different? And which should be used in stream mode?

   await foreach (var completion in completionResult)
            {
                if (completion.Successful)
                {
                    Console.Write(completion.Choices.FirstOrDefault()?.Delta.Content);
                    yield return completion.Choices.FirstOrDefault()?.Message.Content;
                }
// something else
             }

---

support content

The definition of ChatCompletionCreateRequest

I also find this definition in ChatChoiceResponse. So dose it mean we can use whatever Delta or Message in code?