Want to create an interactive transcript for this episode?
Podcast: Arxiv Papers
Episode: Star Attention: Efficient LLM Inference over Long Sequences
Description:
Star Attention enhances Transformer-based LLMs' efficiency by using a block-sparse approximation, reducing memory and inference time by up to 11x while maintaining high accuracy.
https://arxiv.org/abs//2411.17116
YouTube: https://www.youtube.com/@ArxivPapers
TikTok: https://www.tiktok.com/@arxiv_papers
Apple Podcasts: https://podcasts.apple.com/us/podcast/arxiv-papers/id1692476016
Spotify: https://podcasters.spotify.com/pod/show/arxiv-papers