curarin/llm-zoomcamp

Learnngs from 'LLM Bootcamp' (by zoomcamp)

Documentation of my learnings from this course.

• Goal: Put LLMs into production
• Course Link: https://github.com/DataTalksClub/llm-zoomcamp

Preperations

• get .env file and put the following variables:

OPENAI_API_KEY=<your_key>

Week 1:

• What are LLMs? Language Models predict the next token based on previous token. Large Language Models have billions of parameters and are trained on tons of data. It receives an input (usually called a prompt) and returns an output (a prediction).

• What are RAGs? RAG stands for Retrieval Augmented Generation. It is a type of LLM that uses external data sources to improve its performance. Retrieval is search and generation is prediction for output. We put data into a knowledge base / data base, the prompt is sent to the knowledge base, relevant documents are retrieved and sent to the LLM as context.