The paper modifies the DeepMind Chinchilla scaling laws for large language models (LLMs) to include the cost of inference. The analysis suggests that LLM researchers should train smaller and longer models than the Chinchilla-optimal for large inference demand.

https://arxiv.org/abs//2401.00448

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