[Question]: Do you plan to conduct evals?
shippy opened this issue · comments
Simon Podhajsky commented
What is your question?
Hi! Big fan of the concept of coupling piped-content-into-LLMs-with-prompt, but I'm curious if there's any formal process for establishing what version of a prompt works and which doesn't. If not, are you considering creating one, possibly in the format of openai/evals?
Best of luck with everything!