How can I label the inputs to an LLM block in a python transform?

Ask the Community
,
1

Creating some longer LLM inputs in python transforms with 3-4 inputs + some text and labeling. Is there a “correct way” to input this into a Completion Request where I can label what each input is?

I have something like this but am unsure what the ‘ChatMessageRole’ is representing and I want to be able to say “Missing Columns:” + missingColumns within my prompt.

request = GptChatCompletionRequest(
            [ChatMessage(ChatMessageRole.SYSTEM, system_prompt), 
            ChatMessage(ChatMessageRole.USER, parsed_input), 
            ChatMessage(ChatMessageRole.USER, missingColumns),
            ChatMessage(ChatMessageRole.USER, extraColumns)]
        )