Home

Welcome to the prompt_quill wiki!

Here I will try to give you some insight into what Prompt Quill does and how it works.

The basics are Prompt Quill is a simple RAG system, RAG is the short version of retrieval-augmented generation. Retrieval-augmented generation (RAG) is an AI framework for improving the quality of LLM-generated responses by grounding the model on external sources of knowledge to supplement the LLM’s internal representation of information. Implementing RAG in an LLM-based question answering system has two main benefits: It ensures that the model has access to the most current, reliable facts, and that users have access to the model’s sources, ensuring that its claims can be checked for accuracy and ultimately trusted.(found this on a page of IBM)

In our case we don't add facts to the system we add prompts to the system. So when you type in a prompt it will go and find similar prompts in the vector store and out of this it will then generate a prompt for you.

The idea here is to allow to create more sophisticated prompts out of very simple prompts. It helps people like me to create better images from those prompts, as I for myself fail most the time to come up with anything good as a prompt or it takes me long time searching for one or by keep trying very hard to get something working. Most the time this ends up with creating loads of images until I get anywhere near to my idea. Prompt Quill helps in that process by creating a very complex prompt that then is a much better base to work on as my normal starting points like "a nice cat"