I want openai to be able to generate long form content. It does the job well to a certain extent, but it gets random facts from no where and adds it to the text. I have a csv of data that contains the facts.
Can I use the embeddings model to get the relevant data and pass on those facts into the prompt?
Any advice is appreciate. And how should I approach this?
Embedding is the only way you can be sure of the source of information
You will also need a prompt that tells gpt to only use information from the context you provide
To make this work you will need to do embedding and semantic search. And then feed what you find into a final query to gpt.
DM me if you need help
I’m new here! Can’t find the DM button! Can you ‘‘em me instead?