Does Running an AI Model Locally Actually Hurt Its Quality?
Does Running an AI Model Locally Actually Hurt Its Quality?
Posts by Year
2026
sparkDash on the LAN + a Glance widget for it
Deployed sparkDash (multi-DGX-Spark monitoring) to the head Spark node and wired it into Glance. Notes to self so I remember the shape.
Fitting Qwen3.8-27B with 128K Context on a 24GB TITAN RTX: A Self-Hosting Tuning Journey
My agent (Hermes, running on a Mac mini in the house) was “hanging” whenever it talked to my freshly deployed local model server. The server was up, health c...
zotero-smart-ingest: An Agent Skill That Puts Papers in Zotero (With PDFs, No Duplicates)
I read a lot of finance papers. My workflow used to be: find paper → try to download PDF from five different places → manually create a Zotero entry → attach...
给 Hermes Agent 接入火山引擎豆包语音：中文 TTS/STT 插件教程
Hermes Agent 是一个开源的 AI Agent 框架，支持在 Discord、Telegram、iMessage 等平台上运行。默认自带 Edge TTS 和 local whisper，但要获得高质量中文语音，火山引擎的豆包语音是目前的最优解。
2025
Parallel Simulation of Linear Recurrence Relations via Prefix Sum
This post explores a clever algorithm by Guy Blelloch for simulating linear recurrence relations in parallel.
2024
Setting Up Caddy for Reverse Proxy with Cloudflare on Non-Standard Port
In this tutorial, we will walk through the steps to set up Caddy as a reverse proxy server and configure Cloudflare to communicate with your origin server us...
Note to Self: How to work with this site
Local Develop and Preview Install Ruby and Jekyll Follow this post
Journey to the Moon
I am crazy about vim and especially vim binding. However, configuring VIM in the old days of vimscript is just beyond me.