- Python Weekly
- Posts
- Python Weekly (Issue 718 October 16 2025)
Python Weekly (Issue 718 October 16 2025)
Rahul Chaudhary
October 16, 2025
Welcome to issue 718 of Python Weekly. Let's get straight to the links this week.
What 100K+ Engineers Read to Stay Ahead
Your GitHub stars won't save you if you're behind on tech trends.
That's why over 100K engineers read The Code to spot what's coming next.
Get curated tech news, tools, and insights twice a week
Learn about emerging trends you can leverage at work in just 10 mins
Become the engineer who always knows what's next
Articles, Tutorials and Talks
The video teaches how to build MCP servers using the FastMCP Python library. It covers MCP basics, building calculator apps with different communication protocols, integrating APIs for dynamic content, testing with GitHub Copilot, and deploying MCP servers on FastMCP Cloud for a complete development workflow.
This article explains the new features in Python 3.15, compared to 3.14.
The talk explains how to build an MCP server that integrates with ChatGPT using FastMCP and FastAPI. It covers setting up the Python environment, creating and running MCP tools (like a simple add function), and connecting the MCP server to ChatGPT through a public tunnel to enable ChatGPT to call custom functions, enhancing its capabilities with personalized context and tools.
The blog post explains how Celery, which currently does not support native async functions, can still integrate asynchronous Python code by using the async_to_sync utility from asgiref to run async functions synchronously within Celery tasks. It also discusses alternative approaches, such as using a dedicated event loop to run async code inside Celery tasks, and mentions that full async support in Celery is not available yet but is a topic of ongoing development and community solutions.
Turn your ideas into working AI Agents in minutes. Learn how to build, test, and deploy smart tools without writing a single line of code. SPONSOR
The lock file specification for Python, finalized in PEP 751, took more than four years to complete because of the complexity of capturing dependencies across platforms and configurations while maintaining security, readability, and compatibility with different tools. The process required balancing diverse ecosystem needs, resolving dependency graphs, and achieving consensus among major packaging projects, ultimately producing a flexible lock file format called pylock.toml.
This course is designed to guide beginners through the exciting world of Edge AI, covering fundamental concepts, popular models, inference techniques, device-specific applications, model optimization, and the development of intelligent Edge AI agents.
The article describes custom modifications to the Python REPL to make it behave more like a favorite editor, including adding new keyboard shortcuts for code navigation, editing, and inserting example data structures. These changes, enabled through a PYTHONSTARTUP file and packaged in a library called pyrepl-hacks, enhance productivity by allowing quicker code writing and editing with simplified commands and improved usability features in the interactive Python prompt.
The global interpreter lock (GIL) has been interfering with true parallelism in Python. That ends with Python 3.14.
The post presents a straightforward implementation of a neural network from scratch in Python, explaining core concepts such as neurons, layers, weights, biases, activation functions, and training through backpropagation. It demonstrates building and training a simple neural network to approximate a sine wave, highlighting that despite its simplicity, the network can learn complex patterns by adjusting parameters iteratively.
Python Jobs of the Week
Join a Y Combinator startup in New York City developing AI-driven robotics platforms. The role involves Python development across robotics control, FastAPI backends, computer vision, and automation systems for scalable, intelligent hardware.
Interesting Projects, Tools, and Libraries
The best ChatGPT that $100 can buy.
Automatic Video Generation from Scientific Papers.
A terminal-based dashboard for managing cron jobs.
A secure, AI-native IDE for data science.
A command-line tool designed to discover, parse, and test for unauthenticated endpoints using Swagger/OpenAPI documentation. It helps identify potential security issues in unprotected endpoints of APIs, such as PII leaks and common secret exposures.
Django middleware for compressing HTTP responses with Zstandard, Brotli, or Gzip.
Claude Code for CUDA. Free AI assistant that actually understands GPU architecture.
Beautifully designed Reflex components to build your web apps faster. Open source.
New Releases
PyTorch 2.9 introduces new features including a stable libtorch ABI for C++/CUDA extensions, symmetric memory programming for easy multi-GPU kernel development, and enhanced control over graph break handling in torch.compile. It expands wheel support for AMD ROCm, Intel XPU, and CUDA 13, adds FlexAttention optimizations on Intel GPUs and X86 CPUs, and improves Arm platform performance with better backend operators and extended GPU support, aiming to make deep learning development more flexible and efficient.
This release includes the following announcements:
Python Environments extension improvements
Enhanced testing workflow with Copy Test ID in gutter menu
Shell startup improvements for Python environment activation
Upcoming Events and Webinars
There will be following talks
Smarter spreadsheet workflows with Celbridge & Python
Building a "Brain" for a Digital City with Python and LLMs
There will be a workshop, Lightweight Analytics at Scale: Hands-On with DuckDB & Smallpond in Python.
There will be following talks
Imports and Namespaces: A Honking Great idea!
Text classification models with SciKit-Learn
From the Shire: One Test to Rule Them All
There will be a talk, Advanced Safeguarding for AI Inference Pipelines.
There will be following talks
Doing demographic statistics on representation in romantic fanfiction
Orchestrating the complex stuff with Prefect
There will be following talks
Unlocking the Black Box: Demystifying ML Models with Shapley Values
Predicting Extreme Weather Events: Augmenting AI Models to Improve Reliability
Our Other Newsletters |
Programmer Weekly - A free weekly newsletter for programmers.
Founder Weekly - A free weekly newsletter for entrepreneurs featuring best curated content, must read articles, how to guides, tips and tricks, resources, events and more.