EDiT: A Local-SGD-Based Efficient Distributed Training Method for Large Language Models

Add code
Dec 10, 2024

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: