🔒 🧠 We started contributing back to the AI community. We at ConfidentialMind are big believers in open source AI so it makes all the sense for us to contribute back to the community. Our first contribution is an efficiently quantized version of Arcee Supernova Medius SLM by Jaro U. and Edward Kim that we use in some of our AI systems. The model is significantly smaller than its original version, with near similar performance. 2887 downloads in a few days shows that there is interest in models that can be run efficiently on smaller hardware footprint. We make our work fully reproducible to help others innovate faster in this space. Go check out the model here: https://lnkd.in/d3uUT9qE and follow us on Hugging Face 😊
This is awesome. Sharing what you develop and find useful yourselves is a win for everyone!
Markku, Huge congrats to ConfidentialMind for contributing back to the community.
👍🏻👊🏻
Nice work!
Looks great!
Business Expansion Advisor | Strategic Planner | Opportunity Hunter
2wImpressive contribution to the AI community! It's fascinating to see how efficient models can drive innovation. What challenges did your team face during the quantization process? On a different note, I’d be happy to connect—please feel free to send me a request!