Ready to supercharge your Python skills with LLMs? Join me for a hands-on journey through a fun Trip Planning project.

No dull lectures – just hacking, Python code, and pizza breaks which will help you integrate LLMs int your products next day! 🚀🍕

No worries if you're not an expert with all the tools! I have set up Docker Compose for you, and you just need an IDE and some programming skills.

Here's what's in your toolkit:

• Docker Compose
• Python
• Django
• PostgreSQL

Don't stress if you're not a Python pro yet. I have prepped most of the code. You'll learn as we go!

def compute_distance(a, b):
    article_distance = 0
    # article_distance = cosine_similarity(a, b)
    # You just need to uncomment this
    return article_distance

• Tokens
• Embeddings
  • Similarity search
• AI providers
  • OpenAI
  • Google Vertex
• Tracing
• Prompts
• Vector Databases
• RAG

This module dives into the concept of tokens, exploring their significance in LLMs. Participants will explore the mechanisms behind token counting and differences between languages.

Exercise

• Count tokens in different languages
• Find what different tokens represent

Participants will gain a comprehensive understanding of embeddings. The course covers methods for comparing embeddings and leveraging them to identify similarities across documents.

Exercise

• Compute embedding of a text
• Compute distance of two embeddings

This section provides an in-depth comparison between major AI service providers, specifically OpenAI and Google. We will explore the advantages and disadvantages of each, guiding learners in selecting the most appropriate provider for their projects.

Exercise

• Call GPT from OpenAI
• Call Vertex from Google

The module introduces tracing, detailing its applications and benefits in debugging and enhancing AI models. Learners will grasp how tracing can be instrumental in understanding model behavior and performance.

Exercise

• Implement tracing into LangChain

This segment focuses on basic methods of prompting, essential for effectively interacting with AI models. Participants will learn various strategies to elicit desired responses from AI systems.

Exercise

• Write a simple multi shot prompt

Building on the knowledge of embeddings, this module covers the storage and retrieval of embeddings in vector databases. Participants will understand how to efficiently store and search for embeddings to optimize AI applications.

Exercise

• Store embedding in a vector database
• Search for the closest document

As a culmination of the course, participants will apply their acquired skills to develop a real-world project: a Retrieval-Augmented Generation (RAG) based Trip Planner. This project integrates all the concepts learned, showcasing the practical application of LLMs.

Exercise

• Make a Trip planner which uses LLM to prepare a perfect trip to Czech city
• Present your project to other members

I believe that face-to-face interactions are the most effective way to exchange knowledge. This enables individuals to meet and forge connections that can benefit their future endeavors.

You are welcome to join us in a meeting room in Opero Coworking, centrally located in the heart of Prague.

It is also possible to have a workshop in your offices. We will need a meeting room with a projector.

The course fee includes a ready-to-use Git repo with Docker-compose setup, pre-written code, pizza and drinks during the session, and the rental for our meeting space.

All essentials covered for a productive and enjoyable experience.

I'm your tech-savvy, project-polishing, knowledge-sharing enthusiast with real-world LLM magic under my belt.

This isn't your boring lecture, welcome to an electrifying hackathon where creativity meets fun!