corpus.blog
41554 blogs · [ { "id": "01a09547-1598-711b-9e16-a4c038f48955", "title": "Halogen vs Nathanw: Why My Tuned llama.cpp Stack Beats the Hype", "url": "https://sleepingrobots.com/dreams/halogen-vs-nathanw-qwen38-flash-next-strix-halo/", "published_at": "2026-09-11T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea0f5a60b1", "title": "EngramHalo.cpp: Qwen 3.8 Flash Next at 38 t/s on Strix Halo", "url": "https://sleepingrobots.com/dreams/engramhalo-qwen38-flash-next-strix-halo/", "published_at": "2026-08-29T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea0f74033d", "title": "Running Qwen 3.8 Flash Next on Strix Halo: 125B at 20 t/s", "url": "https://sleepingrobots.com/dreams/qwen38-flash-next-strix-halo/", "published_at": "2026-08-29T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea104bc632", "title": "Qwen 3.8 27B on Strix Halo: 18 t/s With MTP, Nearly 2x the Baseline", "url": "https://sleepingrobots.com/dreams/qwen38-mtp-strix-halo/", "published_at": "2026-08-15T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea112fbccf", "title": "Tencent Hy3 on Strix Halo: 10 tok/s Baseline, MTP Doesn't Fit", "url": "https://sleepingrobots.com/dreams/tencent-hy3-iq2m-strix-halo/", "published_at": "2026-07-17T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea1167853d", "title": "Running a 744B Model on Strix Halo with Colibri", "url": "https://sleepingrobots.com/dreams/colibri-744b-moe-strix-halo/", "published_at": "2026-07-16T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea124fab2a", "title": "Benchmarking dots.tts on Strix Halo", "url": "https://sleepingrobots.com/dreams/dots-tts-strix-halo/", "published_at": "2026-06-08T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea1331ad42", "title": "BeeLlama.cpp DFlash on Strix Halo: 2.7x Gemma 31B, But MTP Is Still Faster", "url": "https://sleepingrobots.com/dreams/beellama-dflash-strix-halo/", "published_at": "2026-06-07T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea1352c7a4", "title": "Running the MOSS-TTS Family on Strix Halo", "url": "https://sleepingrobots.com/dreams/moss-tts-strix-halo/", "published_at": "2026-05-30T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea13a5d324", "title": "Meet People Where They Are", "url": "https://sleepingrobots.com/dreams/meet-people-where-they-are/", "published_at": "2026-05-27T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea1463292e", "title": "Testing llama.cpp PR #21344: Faster MoE Prefill, but MTP Fights Back", "url": "https://sleepingrobots.com/dreams/pr21344-moe-kernels-strix-halo/", "published_at": "2026-05-26T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea152c0015", "title": "ROCm 7 on Strix Halo: Benchmarking the New Toolbox Images", "url": "https://sleepingrobots.com/dreams/rocm7-toolbox-upgrade-strix-halo/", "published_at": "2026-05-17T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea15ef5a0d", "title": "Running DramaBox on Strix Halo", "url": "https://sleepingrobots.com/dreams/dramabox-strix-halo/", "published_at": "2026-05-15T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea168e3aed", "title": "Unsloth Studio on Strix Halo: Full GPU Training Without System ROCm", "url": "https://sleepingrobots.com/dreams/unsloth-studio-strix-halo/", "published_at": "2026-05-11T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea16dd4de2", "title": "Gemma 4 MTP Assistant: 3.7x Faster 31B and +45% Faster 26B-A4B on Strix Halo", "url": "https://sleepingrobots.com/dreams/gemma4-mtp-assistant-strix-halo/", "published_at": "2026-05-10T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea16eb15ab", "title": "Optimizing Echo-TTS: CPU Beats GPU", "url": "https://sleepingrobots.com/dreams/echo-tts-optimizations/", "published_at": "2026-05-07T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea17777ec9", "title": "MTP Speculative Decoding: 4.8x Faster Qwen 3.6 27B on Strix Halo", "url": "https://sleepingrobots.com/dreams/mtp-qwen36-strix-halo/", "published_at": "2026-05-06T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea183b48f0", "title": "Benchmarking Echo-TTS on Strix Halo", "url": "https://sleepingrobots.com/dreams/echo-tts-strix-halo/", "published_at": "2026-05-03T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea1882a5af", "title": "Running HY-World 2.0 on Strix Halo: 3D World Reconstruction on an AMD iGPU", "url": "https://sleepingrobots.com/dreams/hy-world-2-strix-halo/", "published_at": "2026-04-17T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea19757ce3", "title": "Friends Don't Let Friends Use Ollama", "url": "https://sleepingrobots.com/dreams/stop-using-ollama/", "published_at": "2026-04-15T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea1a239080", "title": "Speculative Decoding on Strix Halo: 2x Faster Gemma 4 31B Token Generation", "url": "https://sleepingrobots.com/dreams/speculative-decoding-gemma4-strix-halo/", "published_at": "2026-04-12T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea1ae1ab46", "title": "Pi Web UI: A Browser Interface for the Pi Coding Agent", "url": "https://sleepingrobots.com/dreams/pi-web-ui/", "published_at": "2026-04-11T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea1b2e71d9", "title": "Local LLM Infrastructure on Strix Halo", "url": "https://sleepingrobots.com/dreams/local-llm-infrastructure-strix-halo/", "published_at": "2026-04-10T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea1bdff6f2", "title": "Running LLMs on the AMD NPU with Lemonade Server", "url": "https://sleepingrobots.com/dreams/lemonade-server-npu-strix-halo/", "published_at": "2026-04-09T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea1ccbd41a", "title": "Benchmarking OmniVoice on Strix Halo", "url": "https://sleepingrobots.com/dreams/omnivoice-strix-halo/", "published_at": "2026-04-09T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea1dbf44e3", "title": "Benchmarking VoxCPM2 on Strix Halo", "url": "https://sleepingrobots.com/dreams/voxcpm-strix-halo/", "published_at": "2026-04-09T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea1e93e342", "title": "Self-Hosting Fish Audio on Strix Halo", "url": "https://sleepingrobots.com/dreams/fish-audio-strix-halo/", "published_at": "2026-03-15T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea1ed05730", "title": "Medium-Claw: A Persistent AI Companion on Telegram", "url": "https://sleepingrobots.com/dreams/medium-claw/", "published_at": "2026-03-15T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea1f16663a", "title": "LoopMaker Web", "url": "https://sleepingrobots.com/dreams/loopmaker-web/", "published_at": "2026-03-05T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea1f8d5b45", "title": "QuizForge: Self-Learning Quiz Maker", "url": "https://sleepingrobots.com/dreams/self-learning-quiz-maker/", "published_at": "2026-02-20T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea20684409", "title": "Oneiros: A Personal AI Agent Platform", "url": "https://sleepingrobots.com/dreams/oneiros/", "published_at": "2026-01-30T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea20a1c655", "title": "OCR List Maker", "url": "https://sleepingrobots.com/dreams/ocr-list-maker/", "published_at": "2025-12-28T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea2142c0ef", "title": "llama-cpp-python in Docker", "url": "https://sleepingrobots.com/dreams/llama-cpp-python-docker/", "published_at": "2025-11-03T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea218ba0d5", "title": "Lavabo: The Kitchen Sink of Local AI", "url": "https://sleepingrobots.com/dreams/lavabo/", "published_at": "2025-08-10T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea2281efff", "title": "Web Browser Wrapped", "url": "https://sleepingrobots.com/dreams/web-browser-wrapped/", "published_at": "2025-04-14T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea236ee702", "title": "Speech To Text Typing for Wayland Users", "url": "https://sleepingrobots.com/dreams/speech-to-text-typing-wayland/", "published_at": "2025-04-13T00:00:00+00:00" }, { "id": "01a087e2-3504-739e-8123-ceea238c8c60", "title": "Total (Browser) Recall", "url": "https://sleepingrobots.com/dreams/total-browser-recall/", "published_at": "2025-04-13T00:00:00+00:00" }, { "id": "01a087e2-3505-7312-ada1-d6c2bd4a31c0", "title": "A Practical, Fully Local Desktop Voice Agent", "url": "https://sleepingrobots.com/dreams/desktop-voice-agent/", "published_at": "2025-02-05T00:00:00+00:00" }, { "id": "01a087e2-3505-7312-ada1-d6c2bdfe8fff", "title": "A Fully Local, In-Browser Voice Assistant", "url": "https://sleepingrobots.com/dreams/browser-based-voice-assistant/", "published_at": "2025-01-16T00:00:00+00:00" }, { "id": "01a087e2-3505-7312-ada1-d6c2be993bcd", "title": "You Are John", "url": "https://sleepingrobots.com/dreams/you-are-john/", "published_at": "2024-10-13T00:00:00+00:00" } ] posts Claim your blog
Back to sleepingrobots.com
Blog · corpus.blog/blogs/sleepingrobots.com/posts

sleepingrobots.com

sleepingrobots.com

2026

Halogen vs Nathanw: Why My Tuned llama.cpp Stack Beats the Hypeoriginal ↗
11 Sept 2026
EngramHalo.cpp: Qwen 3.8 Flash Next at 38 t/s on Strix Halooriginal ↗
29 Aug 2026
Running Qwen 3.8 Flash Next on Strix Halo: 125B at 20 t/soriginal ↗
29 Aug 2026
Qwen 3.8 27B on Strix Halo: 18 t/s With MTP, Nearly 2x the Baselineoriginal ↗
15 Aug 2026
Tencent Hy3 on Strix Halo: 10 tok/s Baseline, MTP Doesn't Fitoriginal ↗
17 Jul 2026
Running a 744B Model on Strix Halo with Colibrioriginal ↗
16 Jul 2026
Benchmarking dots.tts on Strix Halooriginal ↗
8 Jun 2026
BeeLlama.cpp DFlash on Strix Halo: 2.7x Gemma 31B, But MTP Is Still Fasteroriginal ↗
7 Jun 2026
Running the MOSS-TTS Family on Strix Halooriginal ↗
30 May 2026
Meet People Where They Areoriginal ↗
27 May 2026
Testing llama.cpp PR #21344: Faster MoE Prefill, but MTP Fights Backoriginal ↗
26 May 2026
ROCm 7 on Strix Halo: Benchmarking the New Toolbox Imagesoriginal ↗
17 May 2026
Running DramaBox on Strix Halooriginal ↗
15 May 2026
Unsloth Studio on Strix Halo: Full GPU Training Without System ROCmoriginal ↗
11 May 2026
Gemma 4 MTP Assistant: 3.7x Faster 31B and +45% Faster 26B-A4B on Strix Halooriginal ↗
10 May 2026
Optimizing Echo-TTS: CPU Beats GPUoriginal ↗
7 May 2026
MTP Speculative Decoding: 4.8x Faster Qwen 3.6 27B on Strix Halooriginal ↗
6 May 2026
Benchmarking Echo-TTS on Strix Halooriginal ↗
3 May 2026
Running HY-World 2.0 on Strix Halo: 3D World Reconstruction on an AMD iGPUoriginal ↗
17 Apr 2026
Friends Don't Let Friends Use Ollamaoriginal ↗
15 Apr 2026
Speculative Decoding on Strix Halo: 2x Faster Gemma 4 31B Token Generationoriginal ↗
12 Apr 2026
Pi Web UI: A Browser Interface for the Pi Coding Agentoriginal ↗
11 Apr 2026
Local LLM Infrastructure on Strix Halooriginal ↗
10 Apr 2026
Running LLMs on the AMD NPU with Lemonade Serveroriginal ↗
9 Apr 2026
Benchmarking OmniVoice on Strix Halooriginal ↗
9 Apr 2026
Benchmarking VoxCPM2 on Strix Halooriginal ↗
9 Apr 2026
Self-Hosting Fish Audio on Strix Halooriginal ↗
15 Mar 2026
Medium-Claw: A Persistent AI Companion on Telegramoriginal ↗
15 Mar 2026
LoopMaker Weboriginal ↗
5 Mar 2026
QuizForge: Self-Learning Quiz Makeroriginal ↗
20 Feb 2026
Oneiros: A Personal AI Agent Platformoriginal ↗
30 Jan 2026

2025

OCR List Makeroriginal ↗
28 Dec 2025
llama-cpp-python in Dockeroriginal ↗
3 Nov 2025
Lavabo: The Kitchen Sink of Local AIoriginal ↗
10 Aug 2025
Web Browser Wrappedoriginal ↗
14 Apr 2025
Speech To Text Typing for Wayland Usersoriginal ↗
13 Apr 2025
Total (Browser) Recalloriginal ↗
13 Apr 2025
A Practical, Fully Local Desktop Voice Agentoriginal ↗
5 Feb 2025
A Fully Local, In-Browser Voice Assistantoriginal ↗
16 Jan 2025

2024

You Are Johnoriginal ↗
13 Oct 2024