🏢 University of Exeter
Stable-SPAM: How to Train in 4-Bit More Stably than 16-Bit Adam
·2799 words·14 mins·
loading
·
loading
AI Generated
🤗 Daily Papers
Machine Learning
Deep Learning
🏢 University of Exeter
Stable-SPAM stabilizes 4-bit LLM training, outperforming Adam.