<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Hackathon on Simone Sturniolo's Blog</title><link>https://stur86.github.io/s-plus-plus/tags/hackathon/</link><description>Recent content in Hackathon on Simone Sturniolo's Blog</description><generator>Hugo</generator><language>en-uk</language><lastBuildDate>Sat, 16 May 2026 06:57:08 +0100</lastBuildDate><atom:link href="https://stur86.github.io/s-plus-plus/tags/hackathon/index.xml" rel="self" type="application/rss+xml"/><item><title>Trove - Gemma 4 Hackathon project post-mortem</title><link>https://stur86.github.io/s-plus-plus/posts/trove-post-mortem/</link><pubDate>Sat, 16 May 2026 06:57:08 +0100</pubDate><guid>https://stur86.github.io/s-plus-plus/posts/trove-post-mortem/</guid><description>&lt;p&gt;The &lt;a href="https://www.kaggle.com/competitions/gemma-4-good-hackathon"&gt;Gemma 4 Good Hackathon&lt;/a&gt; is a Kaggle competition that has been sponsored by Google to celebrate the release of their Gemma 4 family of local multimodal LLMs, comprising four models: &lt;code&gt;e2b&lt;/code&gt;, &lt;code&gt;e4b&lt;/code&gt;, each respectively only loading 2 or 4 billion parameters at a time, &lt;code&gt;26b&lt;/code&gt; which is a Mixture-of-Experts model, and &lt;code&gt;31b&lt;/code&gt;. They&amp;rsquo;ve almost immediately been available on &lt;a href="https://ollama.com/"&gt;Ollama&lt;/a&gt; too in quantized form. The theme of the competition was to:&lt;/p&gt;
&lt;blockquote&gt;
&lt;p&gt;[&amp;hellip;] create a solution that addresses a real-world challenge using Gemma 4 models, whether that’s an application that helps millions or a specialized model that could exponentially scale innovation.&lt;/p&gt;</description></item></channel></rss>