Minitron Collection A family of compressed models obtained via pruning and knowledge distillation • 9 items • Updated Oct 3 • 59
Better & Faster Large Language Models via Multi-token Prediction Paper • 2404.19737 • Published Apr 30 • 73