#ml-local · Louche Cannon

mesh.applied | Task-specialized open models for your inference needs

Wed, 14 Oct 2026 12:05:45 +0200

Strix Halo Framework Desktop, 128GB — how long until local AI pays for itself? — Sunk Cost

Tue, 15 Sep 2026 07:31:49 +0200

Bookmarked https://sunkcost.ai/s/framework-desktop-395-128/qwen3.8-27b-q4/?u=11000000&r=15&kwh=0.17&ctx=32768&cs=80&s=smartest

I gave Qwen 3.8 27B a reverse-engineering job I assumed needed a frontier model, and it finished in 30 minutes

Sun, 23 Aug 2026 14:51:36 +0200

Bookmarked https://www.xda-developers.com/qwen-3-8-27b-reverse-engineering-job-frontier-model/

If this is true, the hyperscalers are toast

Thu, 20 Aug 2026 13:26:00 +0200

Bookmarked https://klementoninvesting.substack.com/p/if-this-is-true-the-hyperscalers

GitHub - gavamedia/deltafin: Run Kimi K3, a 2.8T-parameter Mixture-of-Experts LLM, on a single Apple Silicon Mac. Streams MXFP4 experts on demand over HTTP into a local disk cache — fused NEON kernels, Metal/MPS compute, exact reproducible decoding, and

Wed, 29 Jul 2026 07:17:06 +0200

Bookmarked https://github.com/gavamedia/deltafin

Welcome - GAIA SDK

Tue, 14 Apr 2026 07:20:23 +0200

Bookmarked https://amd-gaia.ai/docs

Lemonade: Local AI for Text, Images, and Speech

Thu, 02 Apr 2026 15:03:37 +0200

Bookmarked https://lemonade-server.ai/

My local LLM is the best productivity tool I've installed in years, and it costs nothing to run

Thu, 26 Mar 2026 08:43:50 +0200

Bookmarked https://www.xda-developers.com/local-llm-best-productivity-tool-ive-installed-costs-nothing-to-run/

I access my local AI from anywhere now, and it only took one setting in LM Studio

Tue, 17 Mar 2026 10:01:43 +0200

Bookmarked https://www.xda-developers.com/access-local-ai-from-anywhere-only-took-one-setting-in-lm-studio/

Alibaba's new open source Qwen3.5-Medium models offer Sonnet 4.5 performance on local computers | VentureBeat

Sun, 01 Mar 2026 20:52:48 +0200

Bookmarked https://venturebeat.com/technology/alibabas-new-open-source-qwen3-5-medium-models-offer-sonnet-4-5-performance

Home – Strix Halo HomeLab

Tue, 21 Oct 2025 21:34:51 +0200

Bookmarked https://strixhalo-homelab.d7.wtf/

karpathy/nanochat: The best ChatGPT that $100 can buy.

Sun, 19 Oct 2025 07:40:44 +0200

Bookmarked https://github.com/karpathy/nanochat?tab=readme-ov-file

PaddlePaddle/PaddleOCR-VL · Hugging Face

Thu, 16 Oct 2025 20:56:04 +0200

Bookmarked https://huggingface.co/PaddlePaddle/PaddleOCR-VL

karpathy/nanochat: The best ChatGPT that $100 can buy.

Mon, 13 Oct 2025 18:23:37 +0200

Bookmarked https://github.com/karpathy/nanochat

GitHub - microsoft/edgeai-for-beginners: This course is designed to guide beginners through the exciting world of Edge AI, covering fundamental concepts, popular models, inference techniques, device-specific applications, model optimization, and the devel

Mon, 13 Oct 2025 12:46:29 +0200

Bookmarked https://github.com/microsoft/edgeai-for-beginners

Whispering

Tue, 23 Sep 2025 20:51:24 +0200

Bookmarked https://whispering.bradenwong.com/

Python Audio Transcription: Convert Speech to Text Locally

Tue, 23 Sep 2025 08:01:13 +0200

Bookmarked https://www.pavlinbg.com/posts/python-speech-to-text-guide

Locally AI

Mon, 22 Sep 2025 23:35:29 +0200

Bookmarked https://simonwillison.net/2025/Sep/21/locally-ai/#atom-everything

Experimenting with local LLMs on macOS | Fatih's Personal Blog

Tue, 09 Sep 2025 11:58:16 +0200

Bookmarked https://blog.6nok.org/experimenting-with-local-llms-on-macos/

Setting up local LLMs for R and Python - Posit

Tue, 09 Sep 2025 07:57:30 +0200

Bookmarked https://posit.co/blog/setting-up-local-llms-for-r-and-python/