The Qtile Window Manager: A Python-Powered Tiling Experience
https://tech.stonecharioteer.com/posts/2025/qtile-window-manager/
https://tech.stonecharioteer.com/posts/2025/qtile-window-manager/
Stonecharioteer on Tech
The Qtile Window Manager: A Python-Powered Tiling Experience
My journey from XFCE to Qtile, a tiling window manager written entirely in Python, including setup, configuration strategies, and real-world config examples.
How to use UUIDv7 in Python, Django and PostgreSQL
Learn how to use UUIDv7 today with stable releases of Python 3.14, Django 5.2 and PostgreSQL 18. A step by step guide showing how to generate UUIDv7 in Python, store them in Django models, use PostgreSQL native functions and build time ordered primary keys without writing SQL.
https://www.paulox.net/2025/11/14/how-to-use-uuidv7-in-python-django-and-postgresql/
Learn how to use UUIDv7 today with stable releases of Python 3.14, Django 5.2 and PostgreSQL 18. A step by step guide showing how to generate UUIDv7 in Python, store them in Django models, use PostgreSQL native functions and build time ordered primary keys without writing SQL.
https://www.paulox.net/2025/11/14/how-to-use-uuidv7-in-python-django-and-postgresql/
Paolo Melchiorre
How to use UUIDv7 in Python, Django and PostgreSQL
Learn how to use UUIDv7 today with stable releases of Python 3.14, Django 5.2 and PostgreSQL 18. A step by step guide showing how to generate UUIDv7 in Python, store them in Django models, use PostgreSQL native functions and build time ordered primary keys…
Hachi: An (Image) Search engine
The Hachi project is an end-to-end, fast, self-hosted semantic and metadata search engine designed to enable comprehensive search across all types of media by extracting independent information from distributed personal data. It prioritizes minimal external dependencies, hackability, and the integration of machine learning models to fuse deterministic and semantic attributes, aiming for ...
https://eagledot.xyz/hachi.md.html
The Hachi project is an end-to-end, fast, self-hosted semantic and metadata search engine designed to enable comprehensive search across all types of media by extracting independent information from distributed personal data. It prioritizes minimal external dependencies, hackability, and the integration of machine learning models to fuse deterministic and semantic attributes, aiming for ...
https://eagledot.xyz/hachi.md.html
GNN From Scratch
The article provides an introduction to Graph Neural Networks (GNNs), explaining how graphs are represented for machine learning and introducing the mathematical intuition behind GNNs. It covers key concepts such as nodes, edges, and the message-passing mechanism, helping readers understand how GNNs learn from graph-structured data.
https://cultured-avenue-f13.notion.site/GNN-From-Scratch-2a3dfe9550dd80ac87deee4fe6cd0696
The article provides an introduction to Graph Neural Networks (GNNs), explaining how graphs are represented for machine learning and introducing the mathematical intuition behind GNNs. It covers key concepts such as nodes, edges, and the message-passing mechanism, helping readers understand how GNNs learn from graph-structured data.
https://cultured-avenue-f13.notion.site/GNN-From-Scratch-2a3dfe9550dd80ac87deee4fe6cd0696
cultured-avenue-f13 on Notion
GNN From Scratch | Notion
A blog on decyphering the original GNN paper
Lazy Skills: A Token-Efficient Approach to Dynamic Agent Capabilities
The article presents a method for AI agents to progressively load capabilities on-demand through a three-level system—metadata discovery, detailed documentation loading, and executable tool registration. This approach significantly reduces token usage in large language model contexts, enhances extensibility, and improves efficiency by isolating skills in subprocesses and using keyword-ba...
https://boliv.substack.com/p/lazy-skills-a-token-efficient-approach
The article presents a method for AI agents to progressively load capabilities on-demand through a three-level system—metadata discovery, detailed documentation loading, and executable tool registration. This approach significantly reduces token usage in large language model contexts, enhances extensibility, and improves efficiency by isolating skills in subprocesses and using keyword-ba...
https://boliv.substack.com/p/lazy-skills-a-token-efficient-approach
Substack
Lazy Skills: A Token-Efficient Approach to Dynamic Agent Capabilities
Infinitely scaling CLI agents
Nano Banana can be prompt engineered for extremely nuanced AI image generation
The article highlights Nano Banana’s capability for exceptionally nuanced AI image generation, leveraging a large 32,768-token context window to enable highly detailed and controlled prompts. These features allow for precise adjustments and creative experimentation, pushing the boundaries of prompt engineering in AI art.
https://minimaxir.com/2025/11/nano-banana-prompts
The article highlights Nano Banana’s capability for exceptionally nuanced AI image generation, leveraging a large 32,768-token context window to enable highly detailed and controlled prompts. These features allow for precise adjustments and creative experimentation, pushing the boundaries of prompt engineering in AI art.
https://minimaxir.com/2025/11/nano-banana-prompts
Minimaxir
Nano Banana can be prompt engineered for extremely nuanced AI image generation
Nano Banana allows 32,768 input tokens and I’m going to try to use them all dammit.
RLinf
RLinf is a flexible and scalable open-source infrastructure designed for post-training foundation models (LLMs, VLMs, VLAs) via reinforcement learning.
https://github.com/RLinf/RLinf
RLinf is a flexible and scalable open-source infrastructure designed for post-training foundation models (LLMs, VLMs, VLAs) via reinforcement learning.
https://github.com/RLinf/RLinf
GitHub
GitHub - RLinf/RLinf: RLinf is a flexible and scalable open-source infrastructure designed for post-training foundation models…
RLinf is a flexible and scalable open-source infrastructure designed for post-training foundation models (LLMs, VLMs, VLAs) via reinforcement learning. - RLinf/RLinf