<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Ai on Marc Alier (Ludo)</title><link>https://wasabi.essi.upc.edu/ludo/en/tags/ai/</link><description>Recent content in Ai on Marc Alier (Ludo)</description><generator>Hugo -- gohugo.io</generator><language>en</language><managingEditor>marc.alier@upc.edu (Marc Alier)</managingEditor><webMaster>marc.alier@upc.edu (Marc Alier)</webMaster><copyright>© 2026 Marc Alier</copyright><lastBuildDate>Mon, 11 May 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://wasabi.essi.upc.edu/ludo/en/tags/ai/index.xml" rel="self" type="application/rss+xml"/><item><title>Optimizing My Local LLM Setup for Batch Tasks</title><link>https://wasabi.essi.upc.edu/ludo/en/posts/local-llms-as-bulk-tools/</link><pubDate>Mon, 11 May 2026 00:00:00 +0000</pubDate><author>marc.alier@upc.edu (Marc Alier)</author><guid>https://wasabi.essi.upc.edu/ludo/en/posts/local-llms-as-bulk-tools/</guid><description>&lt;p&gt;In one recent run I had local LLMs going for more than 28 hours, chewing through 437 markdown files. The model used was &lt;code&gt;qwen3.5:122b&lt;/code&gt;, served via Ollama working two Mac Studios on a private LAN. The task was unglamorous: read a file, return a JSON with a two-sentence summary, the named persons mentioned, the topic tags, the key-ideas the file argues. Repeat. Aggregate across files. Write per-entry scaffolds.&lt;/p&gt;</description></item></channel></rss>