Help. Please provide step by step guide for deep learning to generate story

Firstly I’m a beginner with using KNIME. Could somebody kindly provide me a step by step guide for which nodes to use for deep learning to generate and finish my book that I am constantly chopping and changing and rewriting and ultimately struggling to finish.

I would like a detailed guide on which nodes to use and how to use them to achieve the following:
. Feed nodes certain prose (such as Elizabethan).
. Feed nodes my characters (Their names, who they are, what they do).
. Feed nodes the summary of the chapter I want written (Which characters are in the chapter, what I want happen to them in that chapter).

Any assistance would be much appreciated. Thank you.

Hi @JMR1592,
Welcome to the KNIME Forum! We have a team working on integration of Large Language Models (LLMs), but as of now what you are looking for is not really possible in KNIME unless you use one of the community-built ChatGPT integrations. See here for a discussion. But ChatGPT will cost you, while the open-source LLMs I have seen so far have not really lived up to the hype. I do not think they will be able to generate a continuous and coherent story for you.
If you go ahead with ChatGPT/GPT-4, I think you will need to do some experiments yourself. The prose, characters, and summary need to be presented to the LLM in a prompt, a text telling it what to do. This text you give to the model and it should respond with the output. Try it on https://chat.openai.com/chat first and see if the results you get are satisfactory, then you can start automating the generation in KNIME using for example this workflow as a starting point.
Kind regards,
Alexander

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.