The IntelLabs/sqft-phi-3-mini-4k-30-base is a 4 billion parameter sparse language model developed by IntelLabs, derived from Microsoft's Phi-3-mini-4k-instruct. It utilizes the Wanda sparse method to achieve 30% sparsity, focusing on efficient model adaptation. This base model is designed for research into low-cost model adaptation within sparse foundation models.
No reviews yet. Be the first to review!
Hugging Face