Arxiv Papers

[QA] Beyond Cosine Decay: On the effectiveness of Infinite Learning Rate Schedule for Continual Pre-training


Listen Later



This paper compares cosine annealing and infinite learning rate schedules for continual self-supervised learning, finding the latter more effective in enhancing pre-training performance across various datasets.


https://arxiv.org/abs//2503.02844


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


...more
View all episodesView all episodes
Download on the App Store

Arxiv PapersBy Igor Melnyk

  • 5
  • 5
  • 5
  • 5
  • 5

5

3 ratings


More shows like Arxiv Papers

View all
FT News Briefing by Financial Times

FT News Briefing

696 Listeners

Google DeepMind: The Podcast by Hannah Fry

Google DeepMind: The Podcast

199 Listeners

Last Week in AI by Skynet Today

Last Week in AI

289 Listeners

Latent Space: The AI Engineer Podcast by swyx + Alessio

Latent Space: The AI Engineer Podcast

76 Listeners

The AI Daily Brief (Formerly The AI Breakdown): Artificial Intelligence News and Analysis by Nathaniel Whittemore

The AI Daily Brief (Formerly The AI Breakdown): Artificial Intelligence News and Analysis

441 Listeners