PyCon SE 2022

The conference opening by chairwoman Christine Winter.

According to the 2021 State of AI report, "transformers have emerged as a general-purpose architecture for ML. Not just for Natural Language Processing, but also Speech, Computer Vision or even protein structure prediction." Indeed, the Transformer architecture has proven very efficient on a wide variety of Machine Learning tasks. But how can we keep up with the frantic pace of innovation? Do we really need expert skills to leverage these state-of-the-art models? Or is there a shorter path to creating business value in less time? In this code-level talk, we'll show you how to quickly build and deploy machine learning applications based on state-of-the-art Transformers models. Along the way, you'll learn about the portfolio of open source and commercial Hugging Face solutions, and how they can help you deliver high-quality machine learning solutions faster than ever before.

Ok, I lied, I still write tests. But instead of the example-based tests that we normally write, have you heard of property-based testing? By using Hypothesis, instead of thinking about what data I should test it for, it will generate test data, including boundary cases, for you.

Several 5 minutes talks from the conference participants.

"Cron jobs are great! they can save your life and your marriage, they run day and night with any weather and they never miss a shot. But when they become too many it can be a real mess and have you ever tried running a cron job in a container ? The first time I tried Celery I was a bit frustrated, setup is not exactly straight forward and the first time can be really hard but if do not give up the final result is amazing and you will end up with a huge power in your hands."

Behavior Driven Development (BDD) is a testing practice that makes it easy for developers, managers and QA testers to keep track of the key features in a programming project. The talk will provide an introduction to BDD and show to how incorporate it into a project by using `behave`.

Interactive Lectures and Active Learning with Python in Scientific Computing. Teaching programming can be significantly improved if lectures and courses use concepts from active learning. During the last years, I have transitioned from static to interactive slides using Jupyter Notebooks, where students can interact with the material during the lecture. The lectures also contain exercises that the students can perform during the lecture to reinforce their learning process. In some of our libraries, we have also developed interactive functions that enable the students to quickly modify parameters in the computational codes.

"Python for Everyone" is a program I came up to help women from immigrant communities & for people who are from other industry, to move into IT. I will talk about how & what I am organizing as part of PyLadies Stockholm, problems, successes and how to improve.

"Data is becoming one of the most valuable assets of companies and organisations. Needless to say, the importance of being able to work with data is also increasing. To be data literate is about being able to understand and draw insights from data. Python is a key enabler in building the foundation for this. In this session I will go through some different levels of data literacy and how more people can get inspired to work with data and insights with the help of Python."

PyScript is an open source project to bring Python to the browser - the most ubiquitous computing platform in the world. This talk introduces PyScript, explains how it works, the current state of development and what you can expect to see coming in the short, medium and long term.

"Data is an essential part of our work, be it data science, development, UI/product design. It comes in different forms and shapes, represents different concepts and ideas, yet more often than not we stick to limited amount of ways to represent, store, visualize, and query it. Graph databases provide a different perspective on the data that we have, and a way to represent it that puts focus on relationships between entities, highlighting that which might be otherwise overlooked. This can help us to find optimal, simple, and elegant solutions to what might appear to be a hard problem otherwise. In this presentation I would like to share the knowledge and experience of the past few years of me working on applying graph databases to the data and problems, in hope that it will help more people to look at their data in a different way."

Wireless jammer activity from malicious or malfunctioning devices cause significant disruption to mobile network services and user QoE degradation. In practice, detection of such activity is manually intensive and costly, taking days and weeks after the jammer activation to detect it. We present a novel data-driven jammer detection framework termed JADE that leverages continually collected operator-side cell-level KPIs to automate this process.

"With the great momentum in python adoption in industries there is always a dielema on which test framework to choose among pytest and  robot . I would like to share some perspectives of how the  world would look like if these co-existed under one roof .What would a live migration from one to another look like? Effort vs Gain vs Loss.  No verdicts just fun paradoxes!"

Working smarter to fight cancer: using dunder special methods to automate code generation. Anocca has developed a set of novel biotechnologies for deciphering and harnessing the human adaptive immune system to treat cancers. To support our biotech, we have created an information management system and a suite of automated data processing pipelines. The backend is built with Django, Celery, PostgreSQL and a GraphQL API. At Anocca we have a “smarter-not-harder” approach to our development process. We will present the use of the __init_subclass__​ and type​ special methods to automate schema and class generation, reducing boilerplate and repetitive code in our codebase.

Keeping up with the renewable energy transition: Using python and forecasting to maintain a balanced Nordic power system. As the transmission system operator, one of the main tasks of Svenska Kraftnät is to ensure the continuous balance in the power system. This is becoming increasingly complex due to e.g. the increased cooperation between the EU countries, and their energy transition goals. Because of this, Svenska Kraftnät and its Norwegian counterpart is developing a Python based framework for forecasting. By anticipating the imbalance ahead of time, it is possible to automatically procure balancing resources and maintain a stable power system. We will give a background, present the framework and demonstrate some examples and lessons learned from imbalance forecasting.

For most of the workflows involving data and machine learning, it's hard for data professionals to build web applications to communicate their work. Gradio is a library that helps python users easily create UIs for data or machine learning based applications. In this talk, I'll go through how to use Gradio.

"In 2022 birds swarmed our neighborhood. We wanted to chase them away sustainably, kindly, and without using metal spikes. In this talk, we'll learn how to disperse birds safely, with software using Python, AI, and a Raspberry PI."

Discover semantic topics in a collection of documents by leveraging open source projects. "Transformer-based language models have been revolutionizing the field of natural language understanding. In this talk, I will showcase how publicly available language models can be used to discover semantically valid topics from a collection of unstructured texts. This technique is called topic modeling, and can help organize and offer insights to understand large collections of documents. Welcome! "