Jennifer Wei·May 15The Three Horsemen of Numerical Divergence in Hybrid ModelsPart One: Isolating Precision Drift in OLMo Hybrid 7B
InTowards AIbyAdi Insights and Innovations·Jul 5, 2025OLMo 2 vs Claude 3.5 Sonnet: A Head-to-Head AI ShowdownThe AI industry is divided between two powerful philosophies — Open-source democratization and proprietary innovation. OLMo 2(Open Language…
InLevel Up CodingbyMd Monsur ali·Dec 2, 2024OLMo-2–1124: A Production Deployment-Ready, Safest, Efficient, and Reliable LLM ModelHow OLMo 2 7B Instruct ensures reliability, safety, and efficiency for real-world applications in AI-driven solutions
InAi2 BlogbyAi2·Apr 17, 2024OLMo 1.7–7B: A 24 point improvement on MMLUToday, we’ve released an updated version of our 7 billion parameter Open Language Model, OLMo 1.7–7B. This model scores 52 on MMLU, sitting…A response icon1A response icon1
InAi2 BlogbyAi2·Apr 15, 2024Making a switch — Dolma moves to ODC-BYWe’re moving the Dolma dataset to the ODC-BY license. Here’s why.
Inmy_aimlbySanjay Basu, PhD·Feb 7, 2024Deep Dive Analysis of OLMoThe paper “OLMo: Accelerating the Science of Language Models,” authored by 42 researchers, will probably be another seminal article of…
InLevel Up CodingbyAI TutorMaster·Feb 6, 2024OLMo: Democratizing Language Models with Open Access and ArchitectureEmpowering Researchers with Open Access and Architecture: OLMo, the Future of LLM Research
InAi2 BlogbyLuca Soldaini·Aug 18, 2023Dolma: 3 Trillion Token Open Corpus for Language Model PretrainingWe released Dolma, OLMo’s pretraining dataset. Dolma open dataset of 3 trillion tokens. Available on HuggingFace under the ImpACT licenseA response icon1A response icon1
InAi2 BlogbyJen Dumas·Jul 27, 2023The AI2 ImpACT License — A new way to think about AI licensingDeveloped at The Allen Institute for Artificial Intelligence (“AI2”), the AI2 ImpACT License seeks to put AI2’s core organizational values…
