Scalify: scale propagation for efficient low-precision LLM training

Add code
Jul 24, 2024

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: