1 paper
Junseok Lee, Nahun Kim, Sangyong Lee +1
Knowledge distillation (KD) is one of the most effective paradigms for compressing large-scale foundation models into deployable architectures. In the context of Automatic Speech R…