Want to create an interactive transcript for this episode?
Podcast: Arxiv Papers
Episode: Hymba: A Hybrid-head Architecture for Small Language Models