UME: Upcycling Mixture-of-Experts for Scalable and Efficient Automatic Speech Recognition

L Fu, S Yu, S Li, L Fan, Y Wu, X He - arxiv preprint arxiv:2412.17507, 2024 - arxiv.org
Recent advancements in scaling up models have significantly improved performance in
Automatic Speech Recognition (ASR) tasks. However, training large ASR models from …