<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Open Source AI on Osmond van Hemert</title><link>https://www.osmondvanhemert.nl/series/open-source-ai/</link><description>Recent content in Open Source AI on Osmond van Hemert</description><generator>Hugo -- gohugo.io</generator><language>en</language><copyright>© Osmond van Hemert. All rights reserved.</copyright><lastBuildDate>Thu, 14 Aug 2025 00:00:00 +0000</lastBuildDate><atom:link href="https://www.osmondvanhemert.nl/series/open-source-ai/index.xml" rel="self" type="application/rss+xml"/><item><title>Google's Gemma 3 270M — Why Tiny Models Are the Real AI Story</title><link>https://www.osmondvanhemert.nl/posts/250814-gemma3-270m-small-models-big-impact/</link><pubDate>Thu, 14 Aug 2025 00:00:00 +0000</pubDate><guid>https://www.osmondvanhemert.nl/posts/250814-gemma3-270m-small-models-big-impact/</guid><description>Google releases Gemma 3 at 270M parameters, proving that smaller, more efficient models might matter more than the next big model launch.</description></item><item><title>DeepSeek R1 — Open-Source Reasoning Models Change the Game</title><link>https://www.osmondvanhemert.nl/posts/250123-deepseek-r1-open-source-reasoning/</link><pubDate>Thu, 23 Jan 2025 00:00:00 +0000</pubDate><guid>https://www.osmondvanhemert.nl/posts/250123-deepseek-r1-open-source-reasoning/</guid><description>DeepSeek&amp;rsquo;s R1 reasoning model, released as fully open-source with an MIT license, demonstrates that frontier AI capabilities aren&amp;rsquo;t exclusive to US labs anymore.</description></item><item><title>Ollama and the Rise of Local LLMs — Why Running AI on Your Own Hardware Matters</title><link>https://www.osmondvanhemert.nl/posts/240711-ollama-local-llm-revolution/</link><pubDate>Thu, 11 Jul 2024 00:00:00 +0000</pubDate><guid>https://www.osmondvanhemert.nl/posts/240711-ollama-local-llm-revolution/</guid><description>Local LLM tooling has matured rapidly, with Ollama leading the charge. Here&amp;rsquo;s why self-hosted AI is becoming a serious option for developers.</description></item><item><title>Meta Releases Llama 3 — Open Source AI Just Got Serious</title><link>https://www.osmondvanhemert.nl/posts/240418-meta-llama-3-release/</link><pubDate>Thu, 18 Apr 2024 00:00:00 +0000</pubDate><guid>https://www.osmondvanhemert.nl/posts/240418-meta-llama-3-release/</guid><description>Meta&amp;rsquo;s Llama 3 arrives with 8B and 70B parameter models that rival closed-source competitors, reshaping the open-weight AI landscape.</description></item><item><title>Code Llama — Meta's Open Source Bet on AI-Assisted Coding</title><link>https://www.osmondvanhemert.nl/posts/230824-code-llama-open-source-code-generation/</link><pubDate>Thu, 24 Aug 2023 00:00:00 +0000</pubDate><guid>https://www.osmondvanhemert.nl/posts/230824-code-llama-open-source-code-generation/</guid><description>Meta releases Code Llama, a family of open-source code generation models, and it might just change the dynamics of AI-assisted development.</description></item><item><title>Meta Releases Llama 2 — Open Source AI Gets a Massive Boost</title><link>https://www.osmondvanhemert.nl/posts/230720-meta-llama2-open-source-llm/</link><pubDate>Thu, 20 Jul 2023 00:00:00 +0000</pubDate><guid>https://www.osmondvanhemert.nl/posts/230720-meta-llama2-open-source-llm/</guid><description>Meta&amp;rsquo;s release of Llama 2 as a commercially-licensed open model changes the game for developers building with large language models.</description></item><item><title>Meta Releases LLaMA — Open-Source AI Just Got Serious</title><link>https://www.osmondvanhemert.nl/posts/230223-meta-llama-open-source-llm/</link><pubDate>Thu, 23 Feb 2023 00:00:00 +0000</pubDate><guid>https://www.osmondvanhemert.nl/posts/230223-meta-llama-open-source-llm/</guid><description>Meta&amp;rsquo;s release of LLaMA, a family of foundation language models available to researchers, could reshape the AI landscape by democratizing access to powerful LLMs.</description><media:content xmlns:media="http://search.yahoo.com/mrss/" url="https://www.osmondvanhemert.nl/posts/230223-meta-llama-open-source-llm/featured.jpg"/></item><item><title>OpenAI Whisper — Open Source Speech Recognition That Actually Works</title><link>https://www.osmondvanhemert.nl/posts/220922-openai-whisper-speech-recognition/</link><pubDate>Thu, 22 Sep 2022 00:00:00 +0000</pubDate><guid>https://www.osmondvanhemert.nl/posts/220922-openai-whisper-speech-recognition/</guid><description>OpenAI open-sources Whisper, a speech recognition model trained on 680,000 hours of data that approaches human-level accuracy across multiple languages.</description><media:content xmlns:media="http://search.yahoo.com/mrss/" url="https://www.osmondvanhemert.nl/posts/220922-openai-whisper-speech-recognition/featured.jpg"/></item><item><title>Stable Diffusion Goes Open Source — And Changes Everything</title><link>https://www.osmondvanhemert.nl/posts/220908-stable-diffusion-open-source-ai/</link><pubDate>Thu, 08 Sep 2022 00:00:00 +0000</pubDate><guid>https://www.osmondvanhemert.nl/posts/220908-stable-diffusion-open-source-ai/</guid><description>Stability AI&amp;rsquo;s open release of Stable Diffusion marks a watershed moment for generative AI, putting powerful image generation in the hands of every developer.</description><media:content xmlns:media="http://search.yahoo.com/mrss/" url="https://www.osmondvanhemert.nl/posts/220908-stable-diffusion-open-source-ai/featured.jpg"/></item><item><title>Stable Diffusion Goes Public — Open Source AI Image Generation Changes Everything</title><link>https://www.osmondvanhemert.nl/posts/220825-stable-diffusion-open-source-ai-art/</link><pubDate>Thu, 25 Aug 2022 00:00:00 +0000</pubDate><guid>https://www.osmondvanhemert.nl/posts/220825-stable-diffusion-open-source-ai-art/</guid><description>Stability AI releases Stable Diffusion as open source, putting state-of-the-art image generation in the hands of anyone with a GPU. The implications are enormous.</description><media:content xmlns:media="http://search.yahoo.com/mrss/" url="https://www.osmondvanhemert.nl/posts/220825-stable-diffusion-open-source-ai-art/featured.jpg"/></item><item><title>GitHub Copilot and the Open Source Licensing Firestorm</title><link>https://www.osmondvanhemert.nl/posts/210812-github-copilot-open-source-debate/</link><pubDate>Thu, 12 Aug 2021 00:00:00 +0000</pubDate><guid>https://www.osmondvanhemert.nl/posts/210812-github-copilot-open-source-debate/</guid><description>GitHub Copilot&amp;rsquo;s AI-powered code suggestions have sparked a fierce debate about open source licensing, training data consent, and the future of code ownership.</description><media:content xmlns:media="http://search.yahoo.com/mrss/" url="https://www.osmondvanhemert.nl/posts/210812-github-copilot-open-source-debate/featured.jpg"/></item><item><title>AlphaFold's Protein Database — When AI Delivers on the Hype</title><link>https://www.osmondvanhemert.nl/posts/210722-alphafold-protein-database-ai-science/</link><pubDate>Thu, 22 Jul 2021 00:00:00 +0000</pubDate><guid>https://www.osmondvanhemert.nl/posts/210722-alphafold-protein-database-ai-science/</guid><description>DeepMind releases 350,000 protein structure predictions as an open database — a rare moment where AI genuinely accelerates scientific progress.</description><media:content xmlns:media="http://search.yahoo.com/mrss/" url="https://www.osmondvanhemert.nl/posts/210722-alphafold-protein-database-ai-science/featured.jpg"/></item><item><title>AlphaFold 2 — DeepMind Cracks the Protein Folding Problem</title><link>https://www.osmondvanhemert.nl/posts/201126-alphafold2-protein-folding-breakthrough/</link><pubDate>Thu, 26 Nov 2020 00:00:00 +0000</pubDate><guid>https://www.osmondvanhemert.nl/posts/201126-alphafold2-protein-folding-breakthrough/</guid><description>DeepMind&amp;rsquo;s AlphaFold 2 achieves near-experimental accuracy in protein structure prediction at CASP14. This is what AI breakthroughs actually look like.</description><media:content xmlns:media="http://search.yahoo.com/mrss/" url="https://www.osmondvanhemert.nl/posts/201126-alphafold2-protein-folding-breakthrough/featured.jpg"/></item><item><title>Reformer — Can We Make Transformers Practical for the Rest of Us?</title><link>https://www.osmondvanhemert.nl/posts/200123-reformer-efficient-transformers/</link><pubDate>Thu, 23 Jan 2020 00:00:00 +0000</pubDate><guid>https://www.osmondvanhemert.nl/posts/200123-reformer-efficient-transformers/</guid><description>Google&amp;rsquo;s new Reformer model tackles the massive memory and compute costs of Transformers. For engineers building AI-powered features, this matters more than another benchmark score.</description><media:content xmlns:media="http://search.yahoo.com/mrss/" url="https://www.osmondvanhemert.nl/posts/200123-reformer-efficient-transformers/featured.jpg"/></item></channel></rss>