Safely Learning with Private Data: A Federated Learning Framework for Large Language Model

JiaYing Zheng, HaiNan Zhang, LingXiang Wang +3

Private data, being larger and quality-higher than public data, can greatly improve large language models (LLM). However, due to privacy concerns, this data is often dispersed in m…