Hey guys! I work as a QA and I know NOTHING about LLM, and now my company want me to learn how to test our LLM. How can I do that?
You are QA in what field? AI?
Hi, Alex.
Welcome to the DLAI community. It’s a great question! I don’t really have a direct answer, but one reasonable first step would be to scan through the list of all the “short courses” that they have here that discuss various aspects of LLMs. Maybe take a few to get familiar with how to use LLMs and then see if there are any that focus on evaluating the performance of LLMs. E.g. I just saw an announcement of a new course about “Red Teaming” LLMs. “Red Teaming” is not really the same as SQA, but it does involve testing LLMs and evaluating their output, so there might be some insights to be gained there.
The short courses are free at this point and not as much of a time commitment as the specializations like MLS, DLS, M4ML, NLP, GANS and so forth. So that’s a good way to get started and get familiar with LLMs.
Please keep us posted on what you discover about SQA for LLMs! It will obviously be an ongoing discussion with no fixed “end point”.
Thank you for the answer!