August PyMunich Meetup: LLMs, MLFlow, Contextvars for logs

In-person and online: youtube.com/watch?v=Vd-tLlwbM1U

Welcome to our next event on August 29th, hosted by Celus 🤩

Join our youtube channel to stay in the loop.

Featuring three talks. Schedule:

👉 18:20 Doors open, food, drinks, chat

👉 19:15 Warm Welcome by Celus

👉 19:20 Kostiantyn Snihyr: "Kick-off your models versioning with MLflow"

A quick introduction to experiment tracking with MLflow for machine learning projects. The talk covers experiment recording, reproducibility, model registry and briefly touches ML pipelines lifecycle for either beginners or MLOps engineers

Kostiantyn is a Software engineer, who also passionate about machine learning, engineering, and opossums

👉 20:00 Andrey Novikov: "Using Python contextvars to enrich logs with contextual data" 😎

Dive into Python's contextvars and discover how to effortlessly enrich your logs with user actions, resource lifecycles, and more. By setting context information just once, every subsequent log entry in a context (like HTTP requests or task processing) shines with detailed data. No third-party packages needed. Plus, when indexed by Kibana, you gain an unparalleled, comprehensive history of user activities. Join us to unveil a lesser-known use case for contextvars and its transformative logging capabilities

Andrey is a software developer at Celus with a background in mechanical engineering. Experienced in developing cloud-native asyncio python services in startup and large enterprise environments.

👉 20:30 Dibya Chakravorty: "Introducing langsearch: a Python package for harnessing the power of Large Language Models on private data"

Companies usually have a large trove of documentation scattered in various web based systems (internal Wikis, file sharing clouds, etc.). Individuals have a lot of information (text files, code, PowerPoints, PDFs, word docs, photos, etc.) stored in their computers. Wouldn't it be great if we could easily spin up a Large Language Model (LLM) - based personal assistant (like ChatGPT) that knows all about this private data and could easily find answers to your queries? In this talk, I will present a WIP Python package, called Langsearch, that helps you do just this. In this talk, I will talk about the story behind this package, demonstrate the various features, explain how you can start using it, and also touch a little bit on the package's architecture and internals.

Dibya is a Software Engineer at Tech-5 and an experienced testing strategy architect. He is also a known AI enthusiast and organizer of Artificial General Intelligence München meetup group.

👉 21:00: Drinks & Networking

Can't wait to see you!

-Anton, Eugene, Dibya ✌️✌️✌️