AI & ML interests

Efficient Deep Learning

Nyun AI

Welcome to the official Hugging Face page for Nyun AI! We are a deeptech startup focused on building highly optimized AI models to save compute costs, improve deployment speed, and enable deployment on smaller devices. Our mission is to push the boundaries of AI research and development, providing cutting-edge solutions for various industries.

Our Models

nyun-c1-llama3-xxB

  • Description: The nyun-c1 model family are layer pruned llama3 variants. They follow a uniform architecture similar to the original llama3.

nyun-c2-llama3-xxB

  • Description: The nyun-c2 model family have varying number of heads and intermediate size across layers. The largest 61B model outperforms Llama3-70B on a few benchmarks while maintaining fewer parameters, making it an efficient choice for various AI applications.

Our Vision

At Nyun AI, we are dedicated to developing solutions that are agnostic to data and task choices, ensuring that our platform can identify the right set of algorithms for compressing and fine-tuning models. Our goal is to create models that not only perform exceptionally well but also offer practical deployment options for a wide range of devices and applications.

Get in Touch

Interested in collaborating or learning more about our work? Visit our website at www.nyunai.com or reach out to us directly. We are always looking for partners and opportunities to further advance the field of AI.

Latest News

Stay tuned for updates on our latest models and research by following our page here on Hugging Face!

Thank you for visiting our page!


📧 Contact: connect@nyunai.com 🌐 Website: www.nyunai.com