PythonHub
前往频道在 Telegram
2 629
订阅者
-124 小时
+217 天
+5430 天
帖子存档
2 630
Optimizing Tool Selection for LLM Workflows: Differentiable Programming with PyTorch and DSPy
How local, learnable routers can reduce token overhead, lower costs, and bring structure back to agentic workflows.
https://viksit.substack.com/p/optimizing-tool-selection-for-llm?ref=feedle.world
2 630
Types are Transforming Python
The article describes how type annotations and new typing features have transformed Python, making code more robust, maintainable, and easier to understand by enabling static analysis, better tooling, and safer refactoring. Modern additions like data classes, generics, protocols, and TypedDict have improved both developer experience and runtime capabilities, allowing Python to compete mo...
https://spin.atomicobject.com/types-are-transforming-python/
2 630
trae-agent
Trae Agent is an LLM-based agent for general purpose software engineering tasks.
https://github.com/bytedance/trae-agent
2 630
Reflections on 2 years of CPython's JIT Compiler: The good, the bad, the ugly
https://fidget-spinner.github.io/posts/jit-reflections.html
2 630
Solving Wordle with uv's dependency resolver
The post describes how the author models Wordle as a dependency resolution problem, using the Python tool uv to encode Wordle’s feedback and constraints as package dependencies. By representing possible words and letter positions as package versions and leveraging uv’s resolver, the author creates an automated solver that efficiently narrows down solutions based on Wordle guesses and fee...
https://mildbyte.xyz/blog/solving-wordle-with-uv-dependency-resolver/
2 630
Lost Chapter of Automate the Boring Stuff: Audio, Video, and Webcams in Python
https://inventwithpython.com/blog/lost-av-chapter.html
2 630
microsoft / mcp-for-beginners
This open-source curriculum is designed to teach the concepts and fundamentals of the Model Context Protocol (MCP), with practical examples in .NET, Java, TypeScript, JavaScript and Python.
https://github.com/microsoft/mcp-for-beginners
2 630
Scaling recommendations service at OLX
The article details how OLX scaled its Python-based FastAPI recommendation service to handle tens of thousands of requests per second by optimizing infrastructure and architecture. Key strategies included efficient request handling, robust monitoring, and leveraging cloud resources to ensure reliability and scalability during rapid user and data growth.
https://tech.olx.com/scaling-recommendations-service-at-olx-db4548813e3a
2 630
TEN VAD
An open-source solution designed to supercharge your Voice AI Agents with lightning-fast, human-like conversations!Low- latency,lightweight and high performance.
https://github.com/TEN-framework/ten-vad
2 630
Watchfiles: Simple, modern and fast file watching for Python, written in Rust
https://github.com/samuelcolvin/watchfiles
2 630
desto
desto lets you run and manage your bash and Python scripts in the background (inside tmux sessions) through a simple web dashboard. Launch scripts, monitor their and your system's status, view live logs, and control sessions—all from your browser.
https://github.com/kalfasyan/desto
2 630
Agentic checkout in ~100 lines of Python
The post demonstrates how to build a functional agentic checkout system in around 100 lines of Python using LLMs and the Browser Use library, enabling automated online purchases by navigating, filling forms, and extracting order data. It highlights practical challenges like prompt tuning for determinism, implementing structured outputs with Pydantic, and adding guardrails to ensure safe ...
https://sophiabits.com/blog/building-agentic-checkout
2 630
throttled-py: Python rate limiting library with multiple algorithms (Fixed Window, Sliding Window, Token Bucket, Leaky Bucket & GCRA) and storage backends (Redis, In-Memory).
It is available in both synchronous and asynchronous modes, supports waiting and retrying mode, provides access methods such as function call / function calls/decorators/context managers.
https://github.com/ZhuoZhuoCrayon/throttled-py
2 630
Radiate
A fast and flexible evolution engine for implementing artificial evolution and genetic programming techniques.
https://github.com/pkalivas/radiate
2 630
Application Logging in Python: Recipes for Observability
The guide explains how to build robust, production-grade logging in Python by moving beyond basic print statements to use the logging module’s loggers, handlers, formatters, and filters for structured, contextual, and queryable logs. It covers best practices like configuring loggers by module name, using YAML or dictConfig for centralized setup, and adopting JSON output for modern observ...
https://www.dash0.com/guides/logging-in-python
2 630
Python for Nonprofits
Python For Nonprofits demonstrates how to use Python to retrieve, analyze, visualize, and share nonprofit data.
https://github.com/kburchfiel/pfn
2 630
Speed Up Django Queries with values() over only()
Optimizing Django performance? Use values() instead of only() when you only need specific fields, as it skips full model instantiation and avoids triggering deferred field lookups—boosting query efficiency. The article provides practical guidance on when to choose values() for lightweight data access.
https://johnnymetz.com/posts/django-values-over-only/
