<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>QLoRA on Dylan Ler</title><link>http://dylanler.github.io/tags/qlora/</link><description>Recent content in QLoRA on Dylan Ler</description><generator>Hugo -- 0.133.0</generator><language>en-us</language><lastBuildDate>Fri, 06 Feb 2026 08:00:00 -0800</lastBuildDate><atom:link href="http://dylanler.github.io/tags/qlora/index.xml" rel="self" type="application/rss+xml"/><item><title>The Return of ASCII Art: Fine-Tuning a Small LLM to Think in Terminal Diagrams</title><link>http://dylanler.github.io/posts/ascii-tui-diagrams-fine-tuning-qwen3/</link><pubDate>Fri, 06 Feb 2026 08:00:00 -0800</pubDate><guid>http://dylanler.github.io/posts/ascii-tui-diagrams-fine-tuning-qwen3/</guid><description>In an era of photorealistic AI-generated images, I trained a language model to draw with box-drawing characters and pipe symbols.
This isn&amp;rsquo;t nostalgia. It&amp;rsquo;s a bet that the most universal visual medium for AI isn&amp;rsquo;t pixels &amp;ndash; it&amp;rsquo;s text.
Why ASCII Diagrams Still Matter Every developer, every terminal session, every SSH connection, every log file, every README &amp;ndash; text is the one output format that works everywhere. No rendering engine, no GPU, no browser required.</description></item></channel></rss>