G github.com Sinkhorn: Make LLMs even smaller through quantisation while maintaining accuracy 4 points ilitirit 1 year ago 1 comments
