Python Weekly (Issue 627 November 23 2023)

Python Weekly - Issue 627

November 23, 2023

Python Weekly

Welcome to issue 627 of Python Weekly. Let's get straight to the links this week.

From Our Sponsor 

Stepsize integrates with issue trackers like Jira or Linear and analyzes your project data, linking goals and activities to create stunning and digestible sprint and cycle reports. Your first report is completely free, keep all of your stakeholders up-to-date without lifting a finger.

Articles, Tutorials and Talks

Brandt Bucher discusses the development of a Just-In-Time (JIT) compiler for CPython. The talk delves into the challenges and intricacies of implementing a JIT compiler specifically for CPython, the default Python interpreter.

A 12 Lesson course teaching everything you need to know to start building Generative AI applications.

At Meta, Python is integral, powering Instagram's backend, contributing to Python 3.12, and driving crucial aspects like configuration systems and AI work. In the Meta Tech Podcast, Pascal Hartig and Amethyst Reese delve into the Python Foundation Team's efforts, the open-sourced Fixit 2 linter framework, and insights into the role of a production engineer at Meta.

This article tells you more about why these functions are getting the axe, and what to replace them with.

The completion of a project involving the collection, fingerprinting, and indexing of 7 billion small molecules with various structural embeddings, such as MACCS, PubChem, ECFP4, and FCFP4, is announced. The dataset, optimized for molecule search using Unum's USearch, is now globally accessible for free through AWS Open Data, with comprehensive data sheets and visualization scripts available on GitHub.

This post looks into the implementation details of CPython’s Global Interpreter Lock (GIL) and how they changed between Python 3.9 and the current development branch that will become Python 3.13.

Large Language Models (LLMs) can be leveraged for business applications, such as content matching and job search. William Huster demonstrates how to build a prototype application that utilizes LLMs for job search.

How big should your thread pool be? It depends on your use case.

This video explores how generic types in Python 3.12 work, and what the advantage is over just using the Any types.

Local machines can struggle to process large datasets due to memory and network limitations. Coiled Functions provide a cloud-based solution that allows for efficient and cost-effective handling of such extensive datasets, overcoming the constraints of local hardware for complex data processing tasks. Incorporating libraries like Polars can further enhance this approach, leveraging optimized computation capabilities to process data more quickly and efficiently. In this post we’ll use Coiled Functions to process the 150 GB Uber-Lyft dataset on a single cloud machine with Polars.

When writing a Python program, errors are inevitable. However, we can manage the types of errors we produce. Let’s explore a simple model categorizing these errors, from best to worst, and discuss how mindful tool usage can improve software quality.

An overview and some quick examples of using CuDF's Pandas accelerator and how much faster it can be than vanilla Pandas for data analysis.

Building your first neural network could seem like a formidable undertaking, but deep learning frameworks like PyTorch have made the task more accessible than ever. This article explains how to build a neural network using PyTorch.

A step-by-step guide on building Login with Github into your Python apps.

All the technical details of creating a subscription SaaS business using the Python-based Django web framework and Stripe payment processor.

This article discusses four approaches to optimize programs: using a better algorithm, using a better data structure, using a lower-level system, or accepting a less precise solution.

Interesting Projects, Tools and Libraries

Serve 100s of Fine-Tuned LLMs in Production for the Cost of 1.

Config-driven, source control friendly AI application development.

Frigate is an open source NVR built around real-time AI object detection. All processing is performed locally on your own hardware, and your camera feeds never leave your home.

PyNest is a Python framework built on top of FastAPI that follows the modular architecture of NestJS. 

A collection of real world AI/ML exploits for responsibly disclosed vulnerabilities.

pytest-patterns is a plugin for pytest that provides a pattern matching engine optimized for testing.

The best way to use Selenium in Google Colab Notebooks!

Statically typed, purely functional effects for Python.

Automatically turn your SQLalchemy Data Models into a Nice SVG Diagram

Towards Human-Level Text-to-Speech through Style Diffusion and Adversarial Training with Large Speech Language Models

Drop in a screenshot and convert it to clean HTML/Tailwind/JS code.

Neum AI is a best-in-class framework to manage the creation and synchronization of vector embeddings at large scale.

New Releases

Upcoming Events and Webinars

There will be following talks

  • An introduction to Python Metaprogramming

  • Intellectual Property Rights for Data Scientists

  • Leveraging open-source LLMs for production

There will be following talks

  • The CPU in your browser: WebAssembly demystified

  • Four Key Enabling Questions for Agile Delivery

  • Incorporating LLMs into practical NLP workflows

  • Web Hacking: Server Takeover via a Single Python Vulnerability

There will be a talk, Accelerating ML Prototyping: Leveraging HiPlot and Patsy for Feature Engineering at the Speed of Thought.

Our Other Newsletters

 - A free weekly newsletter for programmers.

- A free weekly newsletter for entrepreneurs featuring best curated content, must read articles, how to guides, tips and tricks, resources, events and more.