1 paper · 1 filter
Mohsen Gholami, Mohammad Akbari, Cindy Hu +3
Knowledge distillation from LLMs is essential for the efficient deployment of language models. Prior works have proposed data generation using LLMs for preparing distilled models.…