<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Sleeping Robots</title><description>Project write-ups, experiments, and things I wanted to exist.</description><link>https://sleepingrobots.com/</link><item><title>Pi Web UI: A Browser Interface for the Pi Coding Agent</title><link>https://sleepingrobots.com/dreams/pi-web-ui/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/pi-web-ui/</guid><description>A full-stack web interface that puts the Pi coding agent in the browser — with system-level access, session history, and model switching through a local LiteLLM proxy.</description><pubDate>Sat, 11 Apr 2026 00:00:00 GMT</pubDate><category>agents</category><category>llm</category><category>local</category><category>ai</category><category>infra</category><category>linux</category><author>Zetaphor</author></item><item><title>Local LLM Infrastructure on Strix Halo</title><link>https://sleepingrobots.com/dreams/local-llm-infrastructure-strix-halo/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/local-llm-infrastructure-strix-halo/</guid><description>How LiteLLM, llama-swap, and Lemonade Server compose into a unified local inference platform — routing dozens of models across GPU and NPU through a single API endpoint, accessible anywhere via Tailscale and a local reverse proxy.</description><pubDate>Fri, 10 Apr 2026 00:00:00 GMT</pubDate><category>strix-halo</category><category>llm</category><category>amd</category><category>local</category><category>ai</category><category>hardware</category><category>linux</category><category>infra</category><category>docker</category><author>Zetaphor</author></item><item><title>Running LLMs on the AMD NPU with Lemonade Server</title><link>https://sleepingrobots.com/dreams/lemonade-server-npu-strix-halo/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/lemonade-server-npu-strix-halo/</guid><description>Setting up AMD&apos;s Lemonade Server on Strix Halo to run LLM and Whisper inference on the XDNA 2 NPU — driver builds, architecture decisions, and benchmarks against the integrated GPU.</description><pubDate>Thu, 09 Apr 2026 00:00:00 GMT</pubDate><category>llm</category><category>amd</category><category>local</category><category>ai</category><category>hardware</category><category>linux</category><category>speech-to-text</category><category>infra</category><category>strix-halo</category><author>Zetaphor</author></item><item><title>Benchmarking OmniVoice on Strix Halo</title><link>https://sleepingrobots.com/dreams/omnivoice-strix-halo/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/omnivoice-strix-halo/</guid><description>Running a 600+ language zero-shot TTS model on an AMD integrated GPU — voice cloning benchmarks, ROCm compatibility adventures, and the container workaround that actually worked.</description><pubDate>Thu, 09 Apr 2026 00:00:00 GMT</pubDate><category>tts</category><category>rocm</category><category>amd</category><category>local</category><category>ai</category><category>voice</category><category>strix-halo</category><author>Zetaphor</author></item><item><title>Benchmarking VoxCPM2 on Strix Halo</title><link>https://sleepingrobots.com/dreams/voxcpm-strix-halo/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/voxcpm-strix-halo/</guid><description>Running a 2B parameter tokenizer-free TTS model in both Python and C++ on AMD&apos;s integrated GPU — near-real-time speech synthesis on CPU, and the Vulkan crash that stopped GPU acceleration in its tracks.</description><pubDate>Thu, 09 Apr 2026 00:00:00 GMT</pubDate><category>tts</category><category>amd</category><category>local</category><category>ai</category><category>voice</category><category>linux</category><category>strix-halo</category><author>Zetaphor</author></item><item><title>Self-Hosting Fish Audio on Strix Halo</title><link>https://sleepingrobots.com/dreams/fish-audio-strix-halo/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/fish-audio-strix-halo/</guid><description>Running Fish Audio&apos;s 4B parameter S2-Pro text-to-speech model locally on an AMD Strix Halo integrated GPU via ROCm and Podman.</description><pubDate>Sun, 15 Mar 2026 00:00:00 GMT</pubDate><category>tts</category><category>rocm</category><category>amd</category><category>local</category><category>ai</category><category>strix-halo</category><author>Claude Opus 4.6</author></item><item><title>Medium-Claw: A Persistent AI Companion on Telegram</title><link>https://sleepingrobots.com/dreams/medium-claw/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/medium-claw/</guid><description>A Telegram bot backed by the Pi coding agent with autonomous scheduling, persistent memory, cross-session search, and a web dashboard.</description><pubDate>Sun, 15 Mar 2026 00:00:00 GMT</pubDate><category>agents</category><category>telegram</category><category>local</category><category>ai</category><author>Zetaphor</author></item><item><title>LoopMaker Web</title><link>https://sleepingrobots.com/dreams/loopmaker-web/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/loopmaker-web/</guid><description>A browser-based AI music generation tool powered by ACE-Step, ported to Linux for local generation on AMD Strix Halo hardware.</description><pubDate>Thu, 05 Mar 2026 00:00:00 GMT</pubDate><category>music</category><category>ai</category><category>local</category><category>linux</category><category>strix-halo</category><author>Zetaphor</author></item><item><title>QuizForge: Self-Learning Quiz Maker</title><link>https://sleepingrobots.com/dreams/self-learning-quiz-maker/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/self-learning-quiz-maker/</guid><description>A full-stack quiz platform that turns markdown files and YouTube transcripts into mixed-format quizzes with AI grading, contextual chat, and performance analytics.</description><pubDate>Fri, 20 Feb 2026 00:00:00 GMT</pubDate><category>learning</category><category>ai</category><category>local</category><category>tools</category><author>Zetaphor</author></item><item><title>Oneiros: A Personal AI Agent Platform</title><link>https://sleepingrobots.com/dreams/oneiros/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/oneiros/</guid><description>A modular collection of services for building a personal AI agent — tool use, memory, browser automation, TTS, and multi-platform chat interfaces.</description><pubDate>Fri, 30 Jan 2026 00:00:00 GMT</pubDate><category>agents</category><category>local</category><category>infra</category><category>ai</category><author>Zetaphor</author></item><item><title>OCR List Maker</title><link>https://sleepingrobots.com/dreams/ocr-list-maker/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/ocr-list-maker/</guid><description>Snap a photo of a handwritten list, OCR it with a local vision model, and print a formatted checklist on a thermal receipt printer.</description><pubDate>Sun, 28 Dec 2025 00:00:00 GMT</pubDate><category>ocr</category><category>local</category><category>hardware</category><category>ai</category><author>Zetaphor</author></item><item><title>llama-cpp-python in Docker</title><link>https://sleepingrobots.com/dreams/llama-cpp-python-docker/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/llama-cpp-python-docker/</guid><description>A Dockerfile and docker-compose setup for running llama.cpp with its Python bindings in a container, because finding a working one shouldn&apos;t be this hard.</description><pubDate>Mon, 03 Nov 2025 00:00:00 GMT</pubDate><category>llm</category><category>docker</category><category>local</category><category>infra</category><author>Zetaphor</author></item><item><title>Lavabo: The Kitchen Sink of Local AI</title><link>https://sleepingrobots.com/dreams/lavabo/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/lavabo/</guid><description>An all-in-one Docker container that bundles LLMs, embeddings, vision, and TTS into a single unified inference server.</description><pubDate>Sun, 10 Aug 2025 00:00:00 GMT</pubDate><category>llm</category><category>docker</category><category>local</category><category>infra</category><author>Zetaphor</author></item><item><title>Web Browser Wrapped</title><link>https://sleepingrobots.com/dreams/web-browser-wrapped/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/web-browser-wrapped/</guid><description>Generating weekly Spotify Wrapped-style reports from browser history using local models and Browser Recall data.</description><pubDate>Mon, 14 Apr 2025 00:00:00 GMT</pubDate><category>browser</category><category>recall</category><category>local</category><category>ai</category><author>Zetaphor</author></item><item><title>Speech To Text Typing for Wayland Users</title><link>https://sleepingrobots.com/dreams/speech-to-text-typing-wayland/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/speech-to-text-typing-wayland/</guid><description>Building a custom speech-to-text solution for Linux Wayland users using NVIDIA&apos;s Canary model, Silero VAD, and ydotool.</description><pubDate>Sun, 13 Apr 2025 00:00:00 GMT</pubDate><category>speech-to-text</category><category>linux</category><category>python</category><category>local</category><author>Zetaphor</author></item><item><title>Total (Browser) Recall</title><link>https://sleepingrobots.com/dreams/total-browser-recall/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/total-browser-recall/</guid><description>Building a personal browser history search engine with full-text recall, inspired by Microsoft&apos;s Recall and rewind.ai.</description><pubDate>Sun, 13 Apr 2025 00:00:00 GMT</pubDate><category>browser</category><category>recall</category><category>productivity</category><category>local</category><author>Zetaphor</author></item><item><title>A Practical, Fully Local Desktop Voice Agent</title><link>https://sleepingrobots.com/dreams/desktop-voice-agent/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/desktop-voice-agent/</guid><description>Building a natural language voice controller for the Linux desktop using Qt6, a tiny 1.7B LLM, and a clever vector embedding trick for tool calling.</description><pubDate>Wed, 05 Feb 2025 00:00:00 GMT</pubDate><category>voice</category><category>agent</category><category>local</category><category>linux</category><category>llm</category><author>Zetaphor</author></item><item><title>A Fully Local, In-Browser Voice Assistant</title><link>https://sleepingrobots.com/dreams/browser-based-voice-assistant/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/browser-based-voice-assistant/</guid><description>Building a private, browser-based voice assistant using WebAssembly, Moonshine STT, Piper TTS, and local LLMs.</description><pubDate>Thu, 16 Jan 2025 00:00:00 GMT</pubDate><category>voice</category><category>local</category><category>ai</category><category>webassembly</category><category>llm</category><author>Zetaphor</author></item><item><title>Offline Voice Chatbot in the Browser</title><link>https://sleepingrobots.com/dreams/wasm-voice-chatbot/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/wasm-voice-chatbot/</guid><description>Building a fully offline voice interface for LLMs using WebAssembly — VAD, speech-to-text, and text-to-speech all running client-side.</description><pubDate>Sat, 04 Jan 2025 00:00:00 GMT</pubDate><category>voice</category><category>wasm</category><category>browser</category><category>local</category><author>Zetaphor &amp; Claude Opus 4.6</author></item><item><title>You Are John</title><link>https://sleepingrobots.com/dreams/you-are-john/</link><guid isPermaLink="true">https://sleepingrobots.com/dreams/you-are-john/</guid><description>A text-driven simulation where you interact with a guy named John through natural language, and an LLM determines how his world responds.</description><pubDate>Sun, 13 Oct 2024 00:00:00 GMT</pubDate><category>llm</category><category>games</category><category>simulation</category><category>ai</category><author>Zetaphor</author></item></channel></rss>