Python Weekly (Issue 644 March 28 2024)

Python Weekly - Issue 644

March 28, 2024

Python Weekly

Welcome to issue 644 of Python Weekly. Let's get straight to the links this week.

From Our Sponsor 

A weekly newsletter featuring the best hand curated news, articles, tutorials, talks, tools and libraries etc for programmers.

Articles, Tutorials and Talks

This post explains how Reflex, a pure Python web framework, enables users to build web apps without the need to learn new languages. It details how Reflex works under the hood, highlighting its unique approach of compiling UI to JavaScript while keeping app logic and state management in Python on the server

This post discusses how the author fixed a bug in PyPy's incremental garbage collector that was causing crashes in the AST rewriting phase of pytest on CI environments. It details the technical background on PyPy's incremental GC and the specific issues that were uncovered and resolved in the write barrier implementation.

The article introduces a Python file cache that stores function values in files instead of memory, offering a more persistent caching solution. By using the file_cache decorator, developers can enhance performance by saving time running functions like LLM benchmarks and sharing the cache as a Python module.

The post discusses optimizations made to Django's system check framework, which had gained a reputation for being slow. The optimizations reduced the runtime of running checks on an example client project from 37ms to 18ms, a 50% reduction.

This video guides you through the ins and outs of managing Python virtual environments, while also introducing you to Poetry. You'll learn a host of tips and strategies to manage your projects with greater effectiveness and efficiency.

Tracking a Bad Actor Through the Maze of Obfuscated Malware Code.

Numba can make your numeric code faster, but only if you use it right.

Let's talk about unnecessary else statements in Python.

Looking for data visualization tools for PostgreSQL? We discuss a few options and provide a step-by-step guide on PostgreSQL and Apache Superset.

A dive into the backend side of app development, focusing on building robust APIs with Django and Django Rest Framework.

The video explores LLM architectures, addressing security concerns and critiquing ineffective security tools. It introduces LLM Guard, an open-source tool aiming to enhance LLM security by checking inputs for malicious intent and outputs for sensitive data, demonstrated through practical examples and emphasizing the importance of output monitoring and permissions in safeguarding data.

Predicate pushdown is one of the most important optimizations for a query engine. Read more about it in this article.

The article discusses how Distributed Coroutines, when paired with a distributed scheduler like Dispatch, can simplify the creation of scalable and reliable software by allowing functions to be suspended, serialized, and resumed in another process. It highlights how Distributed Coroutines leverage Python's native support for coroutines and async functions to encode dynamic workflows using regular programming language constructs and control flow.

Learn how to use AWS Lambda to invoke APIs from AWS Bedrock. This video is step by step using Python and boto3's Bedrock client to call the InvokeModel API. It also shows you how to set up your IAM Permissions and modify your Lambda Timeout values to deal with slow responses from Bedrock. 

Interesting Projects, Tools and Libraries

A natural language interface for computers.

Agentic AI Software Engineer. Devika aims to be a competitive open-source alternative to Devin by Cognition AI.

Towards Generic Object Detection via Text-Visual Prompt Synergy.

An open-source project aiming to replicate Devin, an autonomous AI software engineer who is capable of executing complex engineering tasks and collaborating actively with users on software development projects.

Zero-Shot Speech Editing and Text-to-Speech in the Wild.

Make PyTorch models Lightning fast! Thunder is a source to source compiler for PyTorch. It enables using different hardware executors at once.

Provides automated reverse engineering assistance through the use of local large language models (LLMs) on consumer hardware.

Leaping's pytest debugger is a simple, fast and lightweight debugger for Python tests. Leaping traces the execution of your code and allows you to retroactively inspect the state of your program at any time, using an LLM-based debugger with natural language.

The AI-native, open source alternative to Tines / Splunk SOAR. 

RAG Search API.

A Model-Agnostic Framework for Features at Any Resolution.

Upcoming Events and Webinars

There will be following talks

  • Building Retrieval Augmented Generation (RAG) powered applications

  • Getting python out of the way when taking ML models from research to production. A deep dive into the Open Neural Network Exchange (ONNX)

There will be a talk, Spatial Data Visualization with GeoPandas.

There will be following talks

  • Fraud or no Fraud: sounds simple, right?

  • Building GenAI and ML systems with OSS Metaflow

There will be following talks

  • To Bid, or not to Bid – Reinforcement Learning for Real Time Bidding

  • Securing LangChain Applications Against Prompt Injection Using Rebuff

  • Polars is the Pandas killer

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.