PyCon Australia 2025

PyCon Australia 2025 talks videos are available now.

https://www.youtube.com/playlist?list=PLs4CJRBY5F1LRkAAUwbqHNGPBlxDkrz-3

riffq

A toolkit for building PostgreSQL wire-compatible databases in Python, powered by Rust for performance and concurrency.

https://github.com/ybrs/riffq

Tricks from OpenAI gpt-oss YOU ?? can use with transformers

The post details major upgrades that allow models like OpenAI’s GPT-OSS to run, fine-tune, and scale efficiently, including zero-build kernels, 4-bit MXFP4 quantization, tensor and expert parallelism, dynamic layerwise caching, and continuous batching. These improvements cut memory usage, boost speed, and enable larger models to run on affordable hardware, making cutting-edge techniques ...

https://huggingface.co/blog/faster-transformers

VeritasGraph

Enterprise-Grade Graph RAG for Secure, On-Premise AI with Verifiable Attribution.

https://github.com/bibinprathap/VeritasGraph

Avoid Messy Code: Design Patterns for AI Agents in Python

The video demonstrates how to keep Python code for AI agents clean and maintainable by applying design patterns like Chain of Responsibility (for modular pipelines), Observer (for agent logging and context), and Strategy (for pluggable agent personalities). These patterns help decompose logic, improve scalability, and ensure testability for complex AI workflows.

https://www.youtube.com/watch?v=8_liatgLkLc

Hyperparameter Tuning Tips that 99% of Data Scientists Overlook

This video shows how to tune XGBoost models with Optuna while maximizing speed using XGBoost 3.0’s GPU acceleration for 5–15x faster training. He explains why cross-validation is crucial, recommends smart tuning practices, and demonstrates how Optuna’s visualizations help identify impactful hyperparameters in real-world tabular data workflows.

https://www.youtube.com/watch?v=D9xPjkOwpNk

MathFlow

Likerequestsfor mathematical computing, making complex math feel simple.

https://github.com/cybergeek1943/MathFlow

numethods

A lightweight, from-scratch, object-oriented Python package implementing classic numerical methods.

https://github.com/denizd1/numethods

noScribe

Cutting edge AI technology for automated audio transcription. A nice GUI for OpenAIs Whisper and pyannote (speaker identification).

https://github.com/kaixxx/noScribe

Django: Introducing django-watchfiles, for more efficient runserver autoreloading

Django Watchfiles is a library that improves Django's development server by replacing the default autoreloader with the faster, more reliable watchfiles backend. It simplifies setup, enhances reload speed, and brings better cross-platform support with minimal configuration for Django projects.

https://adamj.eu/tech/2025/09/22/introducing-django-watchfiles/

RamTorch

A PyTorch library for memory-efficient deep learning that enables training and inference of large models that don't fit in GPU memory.

https://github.com/lodestone-rock/RamTorch

Wan

Open and Advanced Large-Scale Video Generative Models.

https://github.com/Wan-Video/Wan2.2

Django 6.0 Is Here! CSP Nonces, Background Tasks, Partials & More

The video tutorial covers the new features introduced in Django 6.0 alpha, including built-in Content Security Policy (CSP) nonce support, simpler background task management, and reusable template partials for cleaner code. It provides practical examples and explanations for implementing these features, highlighting improvements in security, asynchronous task handling, and template desig...

https://www.youtube.com/watch?v=doAMlgrTGbE

drf-auth-kit

Modern Django REST Framework authentication toolkit with JWT cookies, social login, and 2FA support.

https://github.com/forthecraft/drf-auth-kit

Klavis

MCP integration layers that let AI agents use thousands of tools reliably.

https://github.com/Klavis-AI/klavis

How I used Cursor AI to migrate a Bash test suite to Python

The migration of a large Bash container test suite to Python using the Cursor AI code editor saved about 1.5 months of development time, with Cursor handling script conversion, function replacement, and automated PyTest suite generation. Although the migration was not entirely smooth and required some manual fixes, the resulting Python test suite passed tests successfully, demonstrating ...

https://developers.redhat.com/articles/2025/09/23/how-i-used-cursor-ai-migrate-bash-test-suite-python#our_real_world_results