Workshop - LLM training with human feedback

Out-of-the-box Large Language Models are great!

But are they also production-level great? To reach the required level we can fine-tune a model on our data for our task. But if you know exactly what requests are handled wrong, can you use this information for training?

Last month Ivan Reznikov did a CodersHQ session on LLMs using several notebooks (see https://github.com/IvanReznikov/DataVerse/tree/main/Courses/LangChain/Lecture2.%20Models) In this workshop in small groups we will do hands on work on the notebooks. Please prepare your laptops, so we can go straight into the matter coming Saturday!