#self-hosting

[raspbian-official]

trying to set up a home server (debian!) so that i can have encrypted storage for multiple users that root can't see and still be able to run other internal services and madness appears to lie this way

[rostglukhov]

Learn how to unload every loaded llama.cpp router model with curl and jq, free VRAM safely, and avoid restarting llama-server in local LLM workflows.

[rostglukhov]

Curated reference of vendor and community inference parameters for Qwen 3.6 and Gemma 4, optimized for agentic workflows and real-world coding systems.

[rostglukhov]

Set up Hermes voice control on Telegram and Discord using local Whisper and free Edge TTS. Includes setup, tuning tips, examples, and troubleshooting.

[rostglukhov]

NemoClaw runs OpenClaw inside OpenShell with policy driven network control, routed inference, and lifecycle tooling. This guide covers quickstart, operations, and troubleshooting.

[rostglukhov]

Compare llama.cpp speeds on a 16 GB GPU for dense and MoE models at 19K, 32K, and 64K context. Tables list VRAM, GPU load, and tokens per second.

[rostglukhov]

Hands-on comparison of LLMs in OpenCode - local Ollama and llama.cpp models vs cloud. Coding tasks, migration map accuracy stats, and honest failure analysis.