MoSLD: An Extremely Parameter-Efficient Mixture-of-Shared LoRAs for Multi-Task Learning.

Lulu Zhao, Weihao Zeng, Xiaofeng Shi, Hua Zhou

Browse the full COLING paper archive.