Want to create an interactive transcript for this episode?
Podcast: Arxiv Papers
Episode: [QA] Cut Your Losses in Large-Vocabulary Language Models
Description:
The paper introduces Cut Cross-Entropy (CCE), a method that significantly reduces memory usage during training of large language models by optimizing cross-entropy loss computation without sacrificing performance.
https://arxiv.org/abs//2411.09009
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