DA-KD: Difficulty-Aware Knowledge Distillation for Efficient Large Language Models

Devs

DA-KD: Difficulty-Aware Knowledge Distillation for Efficient Large Language Models | Read Paper on Bytez