Search – talk-data.com

Title & Speakers	Event
PyData Cluj-Napoca: Meetup #25 2025-03-20 · 16:00 🚀 PyData March Meetup – Join Us! 🐍🎉 Spring is around the corner(just look outside at Feleacu), and so is our next PyData meetup! 🌱📊 Join us this March for another exciting session filled with data, Python, and great conversations. Whether you’re a beginner or an expert, this is the perfect place to learn, share, and connect with fellow data enthusiasts! ---------------------------------------------------------------------------------------------------- "Serialization In The Cloud – Juggling Objects In Processes" by Adrian Mit (30-45min) Most of us have used pickle and basic use cases of serialization in Python. The problem is, there are situations where that just does not cut it. What happens when we have to share objects between jobs in Databricks? What happens if those jobs are spawned from a local machine or simply a Repo that’s unreachable by the target notebook that’s supposed to receive objects?The solution is still serialization, but we need to dive deeper into how it works, when to use which type and why. We’ll explore technologies that offer this solution and how to use them effectively. ---------------------------------------------------------------------------------------------------- "Lightning Talk⚡: Using LLMs Locally – A Practical Guide" by Sebastian Brestin (10-15 min) With the growing availability of large language models (LLMs), it's now possible to run them on moderate hardware, unlocking powerful AI capabilities without relying on cloud services. But how can we integrate these models into our daily workflows? One key advantage of local deployment is enhanced data privacy, allowing users to leverage AI assistance without exposing sensitive information to external servers. In this talk, we'll explore how to set up LLMs on your local machine and use them effectively as coding assistants. ---------------------------------------------------------------------------------------------------- "Wheels, Packages, and Dependencies: A Python Developer's Survival Guide" by Bodgan Bustan (30-45min) This presentation will explore the lifecycle of Python packages - from their internal structure and naming conventions to modern build and dependency management tools. You'll gain an understanding of Python's packaging ecosystem, including the strengths and limitations of traditional tools like pip and conda alongside newer solutions like Poetry and uv. We'll also dive into wheels, their benefits for distribution, and how package repositories fit into the development workflow.Key topics: Package anatomy: Directory structure, metadata, and Python packaging standards Naming conventions and versioning best practices Understanding wheels vs source distributions Comparison of dependency management tools (pip, conda, poetry, uv) Package repositories and distribution strategies Practical workflows for different project needs ---------------------------------------------------------------------------------------------------- NumFOCUS Code of Conduct https://numfocus.org/code-of-conduct	PyData Cluj-Napoca: Meetup #25

🚀 PyData March Meetup – Join Us! 🐍🎉

Spring is around the corner(just look outside at Feleacu), and so is our next PyData meetup! 🌱📊 Join us this March for another exciting session filled with data, Python, and great conversations. Whether you’re a beginner or an expert, this is the perfect place to learn, share, and connect with fellow data enthusiasts!

----------------------------------------------------------------------------------------------------

"Serialization In The Cloud – Juggling Objects In Processes" by Adrian Mit (30-45min)

Most of us have used pickle and basic use cases of serialization in Python. The problem is, there are situations where that just does not cut it. What happens when we have to share objects between jobs in Databricks? What happens if those jobs are spawned from a local machine or simply a Repo that’s unreachable by the target notebook that’s supposed to receive objects?The solution is still serialization, but we need to dive deeper into how it works, when to use which type and why. We’ll explore technologies that offer this solution and how to use them effectively.

----------------------------------------------------------------------------------------------------

"Lightning Talk⚡: Using LLMs Locally – A Practical Guide" by Sebastian Brestin (10-15 min)

With the growing availability of large language models (LLMs), it's now possible to run them on moderate hardware, unlocking powerful AI capabilities without relying on cloud services. But how can we integrate these models into our daily workflows? One key advantage of local deployment is enhanced data privacy, allowing users to leverage AI assistance without exposing sensitive information to external servers. In this talk, we'll explore how to set up LLMs on your local machine and use them effectively as coding assistants.

----------------------------------------------------------------------------------------------------

"Wheels, Packages, and Dependencies: A Python Developer's Survival Guide" by Bodgan Bustan (30-45min)

This presentation will explore the lifecycle of Python packages - from their internal structure and naming conventions to modern build and dependency management tools. You'll gain an understanding of Python's packaging ecosystem, including the strengths and limitations of traditional tools like pip and conda alongside newer solutions like Poetry and uv. We'll also dive into wheels, their benefits for distribution, and how package repositories fit into the development workflow.Key topics:

Package anatomy: Directory structure, metadata, and Python packaging standards
Naming conventions and versioning best practices
Understanding wheels vs source distributions
Comparison of dependency management tools (pip, conda, poetry, uv)
Package repositories and distribution strategies
Practical workflows for different project needs

---------------------------------------------------------------------------------------------------- NumFOCUS Code of Conduct https://numfocus.org/code-of-conduct

Activities & events