LLM's have taken over the world, and their usability has had an unprecedented rate of increase in the last 6 months.

The Alpaca model developed by Stanford's computer science department represents an affordable way to turn pre-trained research LLM's into conversational chatbots that are on par with commercial models.

The project outlines a lot of the key aspects of building a chatbot:

- How to generate and format the data necessary to train everything

- Scripts for training the model with built-in parameters

- Boilerplate code for serving models over Huggingface with Gradio

It's quickly become the new standard for prototyping LLM models.

This talk is designed to give us a hands-on walk through the project, and take a deep dive through its github repo to dissect its most important parts.

This is meant to be a fast-acting introduction to the world of research LLM's, and prompt an open-ended discussion about the current state of language models.

The homepage for the project can be found here: https://crfm.stanford.edu/2023/03/13/alpaca.html

The github repo can be found here: https://github.com/tatsu-lab/stanford_alpaca