Improved Methods for Model Pruning and Knowledge Distillation

Wei Jiang, Anying Fu, Youling Zhang

Model pruning is a performance optimization technique for large language models like R1 or o3-mini. However, existing pruning methods often lead to significant performance degradat…