Articles

How I built an AI-powered system that deploys real Azure infrastructure through natural conversation, running 100% locally with open-source models - no API costs!

Prerequisites Windows with WSL2 installed NVIDIA RTX 5070 GPU NVIDIA drivers installed on Windows host Issue Overview RTX 50-series GPUs use CUDA compute capability sm_120, which requires PyTorch with CUDA 12.8 support. Standard PyTorch installations only support up to sm_90, causing runtime errors. Installation Steps 1. Setup Conda Environment # Install Miniconda (if not already installed) wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh bash Miniconda3-latest-Linux-x86_64.sh source ~/.bashrc # Create new environment conda create -n f5-tts python=3....

Complete installation guide for Fish Speech (OpenAudio) with RTX 50-series GPU support on Windows WSL2. Prerequisites Windows with WSL2 installed NVIDIA RTX 5070 GPU NVIDIA drivers installed on Windows host Miniconda/Anaconda installed Issue Overview RTX 50-series GPUs require PyTorch with CUDA 12.8 support due to their sm_120 compute capability. Standard PyTorch installations cause CUDA kernel errors with newer GPU architectures. Installation Steps 1. Create Conda Environment # Create new environment for Fish Speech conda create -n fish-speech python=3....

A deep dive into creating a robust Docker build system for AFFiNE with custom AI model support, automated version management, and production optimizations.

A complete guide to setting up AFFiNE's AI Copilot feature with custom models, including OpenAI, Anthropic, and experimental Ollama integration.

How I created an automated backup system for Jellyfin using MinIO object storage and .NET 8 to protect against data loss and configuration disasters.

How I hacked my Tuya-based smart plug to avoid a €30/year subscription and built a full self-hosted power usage dashboard.

Building a Scalable Kubernetes Network Architecture with Cilium, BGP, and L2 Announcements Introduction When I began setting up my Kubernetes cluster’s networking infrastructure, I faced a common but challenging problem: how to efficiently manage a limited public IP range while maintaining a clear separation between public and private services. I had a small /29 subnet (X.X.55.89/29) for public services, which meant I needed to be extremely conservative with IP usage. This constraint led me to develop a hybrid networking solution that combines Border Gateway Protocol (BGP) for public services with Layer 2 (L2) announcements for private services....

Preparation Steps Before starting the upgrade, create an etcd backup: talosctl -n 10.122.0.10 etcd snapshot etcd.backup Cluster Configuration My starting point is a Talos cluster with one control plane and four worker nodes: Control Plane: control-01 10.121.0.10 Workers: node01 10.121.0.11 node02 10.121.0.12 node03 10.121.0.13 node04 10.121.0.14 Initial Environment Starting Talos version: 1.6.1 Target Talos version: 1.9.2 Initial Kubernetes version: 1.29.0 Target Kubernetes version: 1.31.5 Hardware: Bare-metal Machine Architecture: amd64 Factory Image Creation Visit https://factory....

This project showcases a sophisticated approach to making personal fitness data searchable and meaningful using modern AI tools. The system combines Strava activity data, vector embeddings, and an intelligent search interface, creating a personalized fitness knowledge base. The system consists of three main components: Strava data collection, vector storage in Supabase, and semantic activity analysis. The workflow begins with authenticating and fetching detailed activity data from Strava using a Python-based OAuth flow....